Hall D Tuesday 13:30-15:30
Anders Niemann1, Samuel Alberg Kock1, Steffen Ringgaard1, Ernst Torben Fründ2, Steffen Ellebæk Petersen, Michael Hasenkam1
1Aarhus University Hospital Skejby, Aarhus N, Denmark; 2Aalborg Hospital, Aarhus University Hospital , Aalborg, Denmark
To generate images suitable for segmentation to be used in CFD simulation of blood flow in fistulas two methods were examined and evaluated. The scans were performed on an in vitro phantom. A mean high steady flow was applied to the inlets of the phantom to simulate the chaotic and turbulent flow in fistulas. Bright blood yielded fast images with excellent image quality except in the fistula chamber, a location with severely turbulent and chaotic flow. Black blood, though slower, yielded excellent image quality everywhere, also in the fistula chamber.
913.
Evaluation of the Thoracic Aorta with Gated CE-MRA: Technical
Feasibility and Comparison with
Ungated Studies
Phillip Young1, Eric Williamson1, Maggie Fung2, David Stanley2, James Glockner1
1Mayo Clinic, Rochester, USA; 2GE Healthcare, Waukesha, Wisconsin, USA
Conventional CE-MRA of the thoracic aorta is limited by motion at the aortic root. We describe our experience with 2 and 4 phase gated CE-MRA sequences, with particular attention to improved visualization of the sinotubular junction, aortic arch vessels, and LCA. Our experience indicates technical feasibility, with markedly improved visualization of the aortic root and proximal coronary arteries resulting from the improved temporal resolution. Prolonged image acquisition and reconstruction times caused limitations in some patients. However, all gated CE-MRA sequences offered improved image quality when compared with prior ungated studies. Gated CE-MRA of the thoracic aorta is a promising technique.
914.
Real Time Self Tracking of Contrast Kinetics for Whole Heart Coronary
Artery Magnetic Resonance
Angiography
Himanshu V. Bhat1, Peng Lai1, Debiao Li1
1Northwestern University, Chicago, Illinois, USA
Contrast enhanced coronary artery magnetic resonance angiography at 3T with slow infusion of contrast agent has recently shown very promising results. In this work a new method for tracking the contrast enhancement during slow infusion of contrast agent is proposed and validated. This method is based on acquiring an extra projection of the heart during imaging and gives an accurate representation of the contrast enhancement. The method is embedded in the high resolution segmented IR-FLASH sequence and has a host of potential applications.
915. Intra-Thoracic Blood Volume Measurement by Contrast Magnetic Resonance Imaging
Massimo Mischi1, Harrie C. M. van den Bosch, Jacques A. den Boer, Jan Verwoerd, Rene` J. Grouls, Cathinka H. Peels, Hendricus H. Korsten
1Eindhoven University of Technology, Eindhoven, Netherlands
The intra-thoracic blood volume (ITBV) is related to the cardiac preload and the left ventricular function. A minimally invasive method for the ITBV measurement by contrast magnetic resonance imaging (MRI) is presented and validated in vitro. The clinical feasibility of the method is also shown. A bolus of a paramagnetic agent is intravenously injected and detected by MRI in the right and left ventricles. The analysis of the measured indicator dilution curves by suitable models, combined with the flow measurement by phase contrast angiography, permits the estimation of the ITBV. The results are accurate and motivate further investigations.
916. 7D Spiral Phase Contrast MRI for the Comprehensive Assessment of Aortic Flow in Mice
Robert L. Janiczek1, Brett R. Blackman1, R. Jack Roy1, Scott T. Acton1, Craig H. Meyer1, Frederick H. Epstein1
1University of Virginia, Charlottesville, Virginia , USA
The ability to assess wall shear stress (WSS) in genetically-engineered mice would enable the investigation of the roles of individual genes in the relationship between WSS and atherosclerosis. The ideal method would cover large regions of the vessel of interest, directly measure WSS, and have high temporal resolution. A 3D stack-of-spirals PC sequence was developed to acquire 7D data of the mouse aorta. High resolution geometry was obtained of the abdominal aorta along with all three-components of velocity throughout systole.
MRA: Continuous Table Movement
Hall D Tuesday 13:30-15:30
971.
Towards Automatic Patient Positioning and Scan Planning Using
Continuously Moving Table Imaging
Peter Koken1, Jochen Keupp1, Sebastian Peter Dries1, Daniel Bystrov1, Peter Börnert1
1Philips Research Europe, Hamburg, Germany
With the increasing number of MRI scan parameters, the operation of a clinical MRI system has become very complex. Improvements in the ease of use of are increasingly important. The idea of this feasibility study is to reduce the operator interaction needed to set up an examination to just selecting the anatomy to be studied by “pushing a single button”. While moving the patient into the MRI magnet, isotropic 3D continuously moving table (CMT) imaging is performed. In parallel, real-time image reconstruction and immediate organ identification is performed using fast image processing. Once the position and extent of the target organ is found, CMT scanning is terminated and the chosen anatomy is automatically positioned in the iso-center. The desired examinations can be started using the automatically derived geometry information without further operator’s interaction.
972.
Continuous Table Movement for Peripheral MRA with Matrix Coils at
3.0T - Comparison to Standard
Step-By-Step MRA
Harald Kramer1, Karin A. Herrmann1, Peter Schmitt2, Michael Zenge2, Christian Glaser1, Maximilian F. Reiser1
1University Hospitals Munich - Grosshadern Campus, Munich, Germany; 2Siemens Medical Solutions, Germany
Because of the well known advantages of magnetic resonance angiography (MRA) like the excellent soft tissue contrast, the lack of ionizing radiation and the possibility of non invasive dynamic imaging, this imaging method established a serious alternative to DSA and CTA. One drawback is the complexity of the exam and the sometimes challenging procedure. One chance to overcome this limitation is the implementation of continuous table movement MRA which shortens and simplifies the entire exam.
973.
Suppression of Image Artifacts Arising from Magnetic Field
Inhomogeneity in Continuous Moving
Table MRI
Yo Taniguchi1, Shinji Kurokawa1, Suguru Yokosawa1, Hisaaki Ochi1, Yoshitaka Bito1
1Hitachi, Ltd., Kokubunji, Japan
Image artifacts and distortions arising from magnetic field inhomogeneity in continuous moving table (CMT) MR images have been investigated using computer simulation. A reconstruction algorithm using phase correction has also been developed from the simulation results, and it was stable in the presence of field inhomogeneity. From the images obtained in phantom experiments it was confirmed that artifacts in the experimental results were similar to those in the simulation results. Furthermore, the artifacts in the images acquired in experiments using both phantoms and volunteers were successfully suppressed by the phase correction technique.
974.
Continuously Moving Table Peripheral CE-MRA (TimCT) on a 1.5 T
Wide-Bore System in an Obese
Population
Florian M. Vogt1, Michael O. Zenge2, Stephan Kannengiesser2, Joerg Barkhausen1, Mark E. Ladd1, Harald H. Quick1
1University Hospital, Essen, Germany; 2Siemens Medical Solutions, Erlangen, Germany
State-of-the-art cylindrical wide-bore MRI scanners provide improved patient comfort for claustrophobic and obese patients. Recently, data acquisition and reconstruction during continuous table movement (TimCT) has been introduced which features 3D coronal slab MRA with centric reordering and increased spatial resolution at the distal end of the large field-of-view. In this study, 5 healthy volunteers and 10 obese patients with known PAOD underwent TimCT CE-MRA. Although examination of the patients on a conventional scanner was impossible because of patient size, moving-table peripheral CE-MRA was successful in all subjects on the wide-bore system. Furthermore, the workflow introduced with TimCT simplified scan planning.
Rita Gouveia Nunes1, Joseph V. Hajnal1, Philip G. Batchelor2, David Atkinson3, David J. Larkman1
1Hammersmith Hospital, Imperial College London, London, UK; 2King's College London, London, UK; 3University College London, London, UK
Continuously moving table approaches allow more efficient acquisition of extended field-of-view images. Unfortunately, as the subject moves through the magnet bore, errors due to imperfections in the B0, gradient and RF fields accumulate leading to severe artefacts. Conventionally, to limit these artefacts, only linear sampling schemes are used and the thickness of the excitation slab reduced. We introduce a generalised method which allows for knowledge of such imperfections to be incorporated into the reconstruction leading to accurate images regardless of the chosen k-space sampling scheme. This flexibility is essential in order to simultaneously acquire multiple images with clinically relevant contrasts.
Hall D Tuesday 13:30-15:30
with T1 Mapping in a Murine Myocardial Infarction Model
Ben Waghorn1, 2, Tiffany Edwards2, Yuhui Yang2, Nathan Yanasak2, Tom Hu2
1Georgia Institute of Technology, Augusta, Georgia, USA; 2Medical College of Georgia, Augusta, Georgia, USA
There is a critical need for non-invasive monitoring of calcium homeostasis in viable myocardial tissue adjacent to necrotic myocardium after myocardial infarction. This study demonstrates that T1 mapping of murine cardiac Manganese-Enhanced MRI can be used to quantify the in-vivo manganese content. Furthermore, the application of this T1 mapping protocol to a myocardial infarction model demonstrates the sensitivity of the technique to delineate regions of the heart with abnormal Mn uptake. This information can potentially be used to estimate salvageable myocardium in a pre-clinical myocardial infarction mouse model.
1021. Optimization of 3-D Tag Sequence and OFM Using a Synthetic Tag Model
Chun Xu1, James J. Pilla1, Gamaliel Isaac2, Aaron Blom1, Joseph H. Gorman1, Robert C. Gorman1, Lawrence Dougherty2
1University of Pennsylvania, Glennolden, Pennsylvania, USA; 2University of Pennsylvania, Philadelphia, Pennsylvania, USA
The aim of this study is to present a novel method of estimating high-resolution 3-D myocardial motion using 3-D tags combined with optical flow method (OFM utilizing a simulation model. A synthetic 3-D tagged cardiac volume was constructed, and deformed by a known systolic flow fields. Prior to phase-to phase pixel displacement estimation, cross correlation coefficient (CC) between the known and estimated flow fields was maximized by adjusting the tag and OFM parameters. This study demonstrates that optimized 3-D OFM combined with 3-D tag sequence has the potential to generate in-vivo myocardial displacement rapidly and accurately.
1022.
Noninvasive Visualization of Myocardial Inflammation Using
Magnetofluorescent Nanoparticle-Contrasted
MRI in Rat Autoimmune Myocarditis
Cheongsoo Park1, Eun Jeong Ahn2, Hyo Eun Park1, Kyuhong Lee1, Tae-Jong Yoon3, Ki-Bae Seung2, Chaejoon Cheong1, Ki Yuk Chang2, Kwan Soo Hong1
1Korea Basic Science Institute, Cheongwon, Republic of Korea; 2the Catholic University, Seoul, Republic of Korea; 3Massachusetts General Hospital, Boston, Massachusetts, USA
Myocarditis is defined as inflammation of the myocardium, mostly caused by viral infection, which leads to autoimmune activation against the host¡¯s own myocardial tissue. Endomyocardial biopsy is still considered to be the gold standard for diagnosing Myocarditis, however it is invasive and low sensitive because of sampling error and high inter-observer variability. Therefore, a novel diagnostic modality to detect the inflammation of myocardium through noninvasive means is needed. Here we investigated whether nanoparticle-contrasted cardiac MRI would be feasible and effective in detecting the status, and discriminating the grade of inflammation in a rat model of experimental autoimmune myocarditis (EAM).
1023.
Cardiac MRI @ 7 Tesla: Initial Experiments in Pigs
Harald H. Quick1, 2, Kai Nassenstein2, Frank Breuckmann2, Stefan Maderwald1, 2, Lena Schäfer2, Mark E. Ladd1, 2, Jörg Barkhausen2
1Erwin L. Hahn Institute for MRI, Essen, Germany; 2University Hospital Essen, Essen, Germany
Cardiac MRI at high field strengths potentially benefits from the increased signal-to-noise ratio (SNR) inherent to high-field MRI. In order to exploit the full SNR potential for cardiac MRI, a number of artifacts and imaging constraints related to the high field strength have to be overcome. The purpose of this study was to perform cardiac MRI in a pig model on a whole-body 7-Tesla MR scanner to evaluate potential advantages and disadvantages specifically associated with cardiac MR imaging at this high field strength.
1024. Cardiac Function in Post-Cardiac Arrest Mice by MRI and Effect of Nitrite Treatment
Stasia Ann Anderson1, Cameron Dezfulian1, Aleksey Alekseyenko1, Mark T. Gladwin1
1National Institutes of Health, Bethesda, Maryland, USA
We examined a mouse model of cardiac arrest by cardiovascular MRI and describe the features of post-arrest cardiac function and the effect of nitrite treatment. We hypothesized that systemic NO2 is depleted during global ischemia (cardiac arrest) and its early repletion could protect the heart from reperfusion injury. Cardiovascular MRI demonstrated the existence and extent of RV dysfunction in a mouse model of cardiac arrest. Systemic nitrite after global ischemia is associated with improved pulmonary blood flow, cardiac function, survival and neurological function in survivors. MRI outcomes indicate RV ejection fraction and contractility are improved by nitrite treatment.
1025. ECG-Gated Cardiac MRI in Mice on a Clinical 3.0T MR Scanner
Jie Huang1, Xiaohai Ma1, 2, Beihua Zhong1, Donna Wang1, Mark DeLano1
1Michigan State University, East Lansing, Michigan, USA; 2Beijing Anzhen Hospital, Beijing, People's Republic of China
Transgenic manipulations in mice are increasingly used to probe genetic and physiological aspects of human cardiovascular physiology. Cardiac MRI in humans is recognized as a robust and accurate method for in vivo assessment of cardiac morphology and function. However, due to the small size and fast rate, cardiac MRI in mice is usually performed on high-field animal scanners significantly limiting the opportunity for cardiac MRI research in mice. In this study we demonstrate the feasibility of performing cardiac MRI in mice with a clinical 3.0T system without the need for an amplifier to detect the R-wave for ECG-gating.
1026.
Cardiovascular Phenotyping of the Mouse Heart Using
4-Dimensional Radial Acquisition and
Liposomal Gd-DTPA
Elizabeth Kathleen Bucholz1, Ketan Ghaghada1, Yi Qi1, Srinivasan Mukundan1, G. Allan Johnson1
1Duke University, Durham, North Carolina, USA
We propose a method for high-throughput cardiovascular phenotyping of the mouse using a 4D radial MRI pulse sequence in conjunction with liposomal Gd-DTPA. To validate the usefulness of the technique, myocardial function was evaluated in a population of 12 mice: 4 C57BL/6J, 4 DBA/2J, and 4 DBA/2J CSQ+. Images were acquired at a resolution of 87x87x348 µm3, with a temporal resolution of 9.6 ms, and 10-12 phases of the heart cycle were captured with a total acquisition time of 16 minutes. Calculation of ejection fraction (EF), end diastolic volume (EDV), and end systolic volume (ESV) were determined to be statistically different for all three populations of mice.
Anna Naumova1, Sarah Fernandes1, 2, Vasily Yarnykh1, Veronica Muskheli1, 2, Chun Yuan1, Charles E. Murry1, 2
1University of Washington, Seattle, Washington, USA; 2Center for Cardiovascular Biology, Institute for Stem Cell and Regenerative Medicine, Seattle, Washington, USA
Cell transplantation using derivatives of embryonic stem cells (ESCs) is a promising therapeutic approach for heart failure. In real clinical practice, the patients with severe heart failure are most in need of cell transplantation therapy. This is the first study to explore restoration effects of hESC-cardiomyocyte transplantation on chronic myocardial infarction. We found that rat cardiac function was moderately improved in one month after cell transplantation into mature infarction. We assume that effect of human cardiomyocyte transplantation to chronic myocardial infarction will become more apparent during a long-term observation, when graft size would reach its functional capability.
1028. An Optical Fiber-Based Gating Device for Cardiac and Abdominal MRI of Small Animals
Adrian Rengle1, Loredana Baboi1, Hervé Saint-Jalmes2, 3, Raphaël Sablong1, Olivier Beuf1
1INSA-Lyon, Université Lyon 1, Villeurbanne, France; 2Faculté de Médecine, Université Rennes 1, Rennes, France; 3Centre Eugène Marquis, Rennes, France
An optical-based device designed to synchronize MRI acquisition on small animals was developed using a transmit-receive pair of optical fibers. Light from a laser diode was focused into the transmit fiber and impinged upon the moving skin. The reflected light was detected by the receive fiber and then carried to a light-voltage amplified photodiode. The output signal was interconnected with a commercial trigger unit. The optical-based signals recorded on mice were correlated with both respiratory and heart motions. Signal amplitudes were large enough to perform an easy adjustment of gating level with good differentiation between cardiac and respiratory signal. The signal was totally unaffected by radiofrequency pulses or currents induced by the magnetic field gradients switching used for imaging. This optical-based gating device was used successfully for dual cardiac and respiratory synchronization for heart and liver examinations of mice at 4.7T. The device developed using thin optical fibers is simple to use and well suitable for small animal MRI using high field strength narrow bore systems.
1029. Cine-MRI vs. 2D-Echocardiography to Measure Left Ventricular Function in Rat Heart in Vivo
Daniel J. Stuckey1, Carolyn A. Carr1, Damian J. Tyler1, Kieran Clarke1
1University of Oxford, Oxford, UK
Two dimensional echocardiography is the most commonly used method for studying cardiac morphology and function in small animals. We have compared 2D-echo with cine-MRI measurements of function in control and infarcted rat hearts and demonstrated strong correlations between the two modalities. However, cine-MRI had greater reproducibility and left ventricular ejection fractions were 12 „b 6% higher when measured using MRI. The accuracy of cine-MRI allows the identification of alterations in heart function that may be missed if using 2D-echo. Therefore, caution should be taken when comparing functional results acquired using short axis 2D-echo vs. cine-MRI.
Arunark Kolipaka1, Kiaran P. McGee1, Anthony J. Romano2, Kevin J. Glaser1, Philip A. Araoz1, Armando Manduca1, Richard L. Ehman1
1Mayo Clinic, Rochester, USA; 2Naval Research Laboratory, Washington, District Of Columbia, USA
Knowledge of the mechanical properties of the myocardium has the potential to provide clinically relevant data for diagnosing a variety of cardiac disease processes. MR elastography (MRE) is a phase-contrast MR-based method for spatially resolving shear stiffness. We propose a new MRE-based model for estimating shear stiffness in a spherical shell geometry using a model of shear wave propagation within a thin spherical shell. This study describes the analysis of MRE-derived estimates of shear modulus using this model in varying thickness shells. This model may find useful applications in the heart, eye and bladder.
Arunark Kolipaka1, Kiaran P. McGee1, Philip A. Araoz1, Richard L. Ehman1
1Mayo Clinic, Rochester, USA
It is appreciated that change to the mechanical properties of the myocardium are associated with a variety of cardiac disease processes. MR elastography (MRE) is a phase-contrast based MR technique capable of quantitating shear stiffness under static imaging conditions. This study describes the use of MRE to measure shear stiffness in a phantom undergoing dynamic volumetric changes. MRE derived measures of shear stiffness are compared to those derived from a pressure-volume derived method for calculating shear modulus of the left ventricle.
Hall D Tuesday 13:30-15:30
of Items/Devices Used in MR Environments
Gregor Schaefers1
1MR:comp GmbH, Gelsenkirchen, Germany
MR safety and image compatibility are internationally recognized as important issues for medical devices. Medical devices and items that can be exposed to an MR environment must be tested on magnetically induced forces, torques, RF heating, induction of voltages and safe functioning as well as MR image artifacts. First standardized test methods were already established. Further standard development is necessary in order to minimize patient risk and guiding device manufacturers in development of MR safe devices as well as supporting the MR user with meaningful experimental results.
1057. Material- And B0-Dependent Scaling of Torque Effects
Roger Luechinger1, Volkert A. Zeijlemaker2, Maarten van Bentem2, Firat Duru3, Peter Boesiger1
1Institute for Biomedical Engineering, University and ETH, Zurich, Switzerland; 2Medtronic Inc., Netherlands; 3University Hospital, Zurich, Switzerland
The scaling with the main magnetic field strength of torque effects on different materials (NdFeB magnet, nickel plate, and Dysprosium sulfate) and on a medical device (Reveal loop recorder) have been investigated. The measurements have been performed at 0.15T, 0.5T, 1T, 1.5T, 3T, and 7T. For ferromagnetic material maximal torque will not change, above its saturation point. Saturation will occur in most ferromagnetic materials in the range of 0.2-1T. Torque on magnets will increase linearly with the main magnetic field strength.
1058. Torque Measurements in MRI Safety Testing
Nikolaus M. Szeverenyi1, Mugdha Thakur2
1SUNY Upstate Medical University, Syracuse, New York, USA; 2Syracuse University, Syracuse, New York, USA
ASMT standard F2213-04 specifies how torque measurements are to be obtained in static magnetic fields using a torque balance, but there exists very little literature from investigators actually using such a device. Instead approximations involving rotation of the ferro-metallic sample on a plastic surface or other qualitative observations are usually employed. We describe the fabrication and use of several modified versions of the ASTM torque balance that will be appreciated by the MRI safety testing community. The origins of this force and its field dependence are also investigated.
1059.
An EEG System with Carbon Wire Electrodes and an
Anti-Polarization Circuit for Simultaneous
EEG-FMRI Recording
Michiro Negishi1, Ilan Laufer1, Mark Abildgaard1, Terry Nixon1, Robert Todd Constable1
1Yale University, New Haven, Connecticut, USA
Simultaneous EEG-fMRI (Electroencephalography-functional Magnetic Resonance Imaging) recording offers high temporal resolution electrophysiological recording and high spatial resolution hemodynamic recording from the same experimental runs. Carbon wire electrodes (not solid electrodes with carbon leads) are suitable for simultaneous EEG-fMRI recording because they cause less radio frequency heating and less susceptibility artifacts than metallic electrodes. However, carbon wire electrodes have not been used widely in human EEG because of the electrode polarization, or imbalance of DC potentials among electrodes. In this study, we developed and evaluated a prototype EEG system with carbon wire electrodes and a pre-amplifier equipped with an anti-polarization mechanism.
1060.
Safety of Localising Intracranial EEG Electrodes Using MRI: A
Comparison Between Head and
Body Coils at 3T
David William Carmichael1, John S. Thornton2, Philip J. Allen2, Louis Lemieux1
1UCL Institute of Neurology, London, UK; 2National Hospital for Neurology and Neurosurgery, London, UK
The safety of post-implantation localisation of intracranial EEG electrodes by MRI was investigated for body and head coils at 3T using a test object with a combination of electrodes simulating a clinical arrangement and a high SAR sequence. For head coil RF-transmission moderate heating was observed (<2ºC) under ‘standard’ conditions, with the external electrode leads (tails) separated, increasing with the tails in electrical contact. Conversely, for body coil RF–transmission, heating was markedly higher (+6.4ºC) with the tails separated. MRI with intracranial EEG electrodes at 3T can be safe, providing SAR is restricted and a head transmit coil is used.
1061.
Estimating Specific Absorption Rate (SAR) During MRI in the
Human Brain with Intracranial EEG
Electrodes Used for Epilepsy
Monitoring: A Preliminary Study Using Finite Integral Technique (FIT)
Modelling
David William Carmichael1, Yan Li2, Andrew McEvoy3, Jeff W. Hand2, 4, Louis Lemieux1
1UCL Institute of Neurology, London, UK; 2Imperial College, London, UK; 3National Hospital for Neurology and Neurosurgery, London, UK; 4Imperial College Healthcare NHS Trust, London, UK
A commercial EM solver was used to model the interaction between an MRI head coil and the human head with a subdural-grid implant used for epilepsy monitoring. The electric field was found to be focused around the implant, with much greater values than seen in the head without the implant. However, when SAR was averaged over both 1g and 10g masses the peak local SAR proximal to the implant increased but the position of maximum local SAR produced (in the sinus region away from the implant) was not altered.
1062. Heating Effects Measured in EEG Electrodes at 3T
Ruth L. O'Gorman1, 2, Laura A. Wherity3, Sophie F. Riches4, Owen G. O'Daly2, Dominic H. ffytche2
1King's College Hospital, London, UK; 2Institute of Psychiatry, London, UK; 3King's College Hospital, UK; 4Institute of Cancer Research, London, UK
The simultaneous acquisition of EEG and fMRI data provides a unique opportunity for investigating cerebral function at high spatial and temporal resolution, but simultaneous EEG-MRI raises several patient safety issues. This study investigated the heating effects from several MRI pulse sequences in EEG electrodes placed in a variety of locations and orientations. A maximum temperature rise of 1.44 ºC was recorded, but the time course of the temperature measurements suggests that greater temperature rises may be likely if high-SAR sequences are applied consecutively. In addition, a large variation in recorded temperature change was observed across different electrode positions and orientations.
1063. Improved RF Safety of Interventional Devices Using Cable Traps
Krishna N. Kurpad1, Erik T. Bieging1, Orhan Unal1, 2
1University of Wisconsin , Madison, USA; 2University of Wisconsin - Madison, Madison, USA
RF safety is a major concern in the design of active tracking and imaging devices for MR guided interventions. In this work, we demonstrate the potential of simple coaxial cable traps in suppressing RF heating to acceptable levels.
1064. Reduction of RF Heating of Interventional Cryoprobes Using Chokes
Sonal Josan1, Ronald Watkins1, Bruce Daniel1, Kim Butts Pauly1
1Stanford University, Stanford, California , USA
In presence of electrically conductive structures, such as implants, wires, or interventional devices, radio-frequency fields used in MRI can cause significant heating of the surrounding tissue. The magnitude of the heat depends on the device geometry, position within the patient, and position relative to the RF coil electric field. These effects have been described for guidewires at 1.5T and higher. The purpose of this work is to show resonant RF heating from a cryoablation device at 0.5T, to determine the safe & worst-case configurations, and demonstrate the use of RF chokes to reduce unwanted currents on the cable that lead to tissue heating.
Hall D Tuesday 13:30-15:30
Kai Kratt1, Ulrike Wallrabe1, Jan G. Korvink1
1University of Freiburg - IMTEK, Freiburg, Germany
We present the development of 3 D solenoidal micro coils for MR analysis based on an automatic wire bonder. By developing a stable and repeatable bond process with insulated wire, micro coils with sub millimeter diameter have been manufactured. The winding process for a single coil takes about 200 ms, whereas the manufacture of a 100 coil array takes less than a minute. Micro coils with 4 windings and inner diameter of 200 µm exhibit an inductance of 12.7 nH and a resistance of 580 mW at 300 MHz. A quality factor of 41 enables high-resolution MRI/NMR.
1099. An Inductively Decoupled Coil Array for Parallel Imaging of Small Animals at 7T
George Carlos do Nascimento1, Fernando Fernandes Paiva1, Afonso C. Silva1
1National Institutes of Health, Bethesda, USA
A two channel small animal coil array for 7T was designed based upon the method of inductive decoupling between the channels. The level ofisolation can be adjusted by adjusting the coupling in the passive transformer.The method shows a very nice immunity to standing waves, cross talking effects and otherparasitic signals in the array channels. The acquired images shows also that the method provides a good signal to noise ratio.
1100. A 4-Channel Transceive Surface Coil Array for Small Animal Imaging at 9.4T
Samuel O. Oduneye1, 2, Ravi S. Menon1, 2
1The University of Western Ontario, London, Canada; 2Robarts Research Institute, London, Canada
As operational frequencies increase linearly with higher static fields, the wavelength approaches the size of the sample being imaged. The resulting standing wave mode deteriorates image homogeneity. With phased array surface coils, the produced B1 field can be tailored to overcome the so called ‘dielectric resonance effect’, high RF power deposition and signal radiation losses. Here we present a novel high field transceive surface coil array for small animal imaging at 9.4T. Additionally, this design allows the coil to be employed for fast parallel imaging techniques while maintaining the high signal to noise ratio inherent advantage of surface coil designs.
Stefan Fischer1, Florian Martin Meise1, Andrea Kronfeld1, Beat Alessandri1, Wolfgang G. Schreiber1
1Mainz University Medical School, Mainz, Germany
Animal models with small animals like rats and mice are common in the research on diseases and its treatment. For those studies the simultaneous examination of multiple small animals is preferable to avoid an overall long scan time for large numbers of animals. A design without shielding, resulting in an inexpensive and simple array construction, allows time effective building of individual coil setups. Theoretically it can be expanded with further elements for multi channel receive systems up to 32 channels, if state-of-the-art scanner technology is used.
1102. Four Channel Array for 9.4T Animal Studies
Christopher Joseph Wargo1, John C. Gore1, Malcolm J. Avison1
1Vanderbilt University, Nashville, Tennessee, USA
Parallel arrays are commonly used in human clinical MR studies due to the benefits they provide. Beyond SNR improvement compared to a volume coil, parallel arrays support rapid imaging methods such as SENSE. The reduced scan time can be used to improve experiment efficiency, increase resolution, or help reduce field-dependent T2* and B0 inhomogeneity blurring and distortion artifacts. To date, parallel imaging has not been widely adapted to animal scanner applications, where advantages for improved performance are sometimes less obvious. We have developed a four channel array intended for 9.4T animal imaging, and evaluated parallel imaging performance in rat brain.
1103. Optimization of Phased Array Coils for Small-Animal MRI at 9.4T
Zhangwei Wang1, Martin Tabbert2, Sven Junge2, Roy E. Gordon2, Qing X. Yang1, Michael B. Smith3, Christopher M. Collins1
1The Pennsylvania State University, Hershey, Pennsylvania, USA; 2BRUKER Biospin GmbH, Ettlingen, Germany; 3Novartis Institutes for BioMedical Research, Inc., Boston, Massachusetts, USA
Animal imaging with phased array coils in high field MRI has become increasingly routine. Mice and rats are among the species most often imaged. To increase the SNR in comparison to existing standard quadrature birdcage coils and an existing phased array coil design, we have performed a numerical analysis and comparison of several proposed array geometries loaded with either a rat head or a mouse body considering the axial B1-homogeneity, B1-sensitivity and g-factor distribution.
1104.
Design of a Highly Sensitive Solenoid-Based RF-Coil for Small
Animal Brain Imaging
Hisaaki OCHI1, Satoshi Minoshima1, Donna J. Cross1, Cecil E. Hayes1
1University of Washington, Seattle, Washington, USA
A highly sensitive solenoid-based RF-coil whose sensitivity distribution is suitable for small animal brain imaging has been designed and fabricated. The designed coil was constructed as a combination of solenoid, Counter-Rotating-Current (CRC), and surface coils. We made a current loop at the top of the CRC coil so that the coupling between CRC and surface coils can be canceled out by adjusting the area of the top loop of the CRC coil. The sensitivity of this coil is 30% or more greater than those of QD birdcage coil and multi-channel phased-array coil of the same size.
1105. 20-Channel Mouse Phased-Array Coil for Clinical 3 Tesla MRI Scanner
Boris Keil1, Lawrence L. Wald2, Graham C. Wiggins2, Christina Triantafyllou3, Florian M. Meise4, Klaus Jochen Klose1, Johannes T. Heverhagen1
1Philipps University, Marburg, Germany; 2A.A. Martinos Center for Biomedical Imaging, MGH Dept. of Radiology, Harvard Medical School, Charlestown, Massachusetts, USA; 3A.A. Martinos Center at McGovern Institute for Brain Research, MIT, Cambridge, Massachusetts, USA; 4Mainz University School, Germany
The purpose of this study was to develop a dedicated 20-channel phased-array coil for mice imaging using a clinical 3 Tesla MRI system. Especially the challenges for the miniature design and the construction of large numbers phased-array mouse coils is presented. A tiny tubular conductors with an inner and outer diameter of 0.8 mm and 1.2 mm respectively was chosen to build all 20 coil elements. The preamplifiers impedance was transformed to a short at the detuning circuits to provide preamplifier decoupling. The small coil geometry shows good decoupling between elements and performs well in SNR.
1106. Multiple-Mouse MRI with Multiple Arrays of Receive Coils (MARCs)
Marc Stephen Ramirez1, James Andrew Bankson1
1The University of Texas M. D. Anderson Cancer Center, Houston, Texas, USA
The efficiency of small-animal MRI studies can be improved by simultaneously scanning multiple animals within a single MRI scanner. To date, this has involved dedicating up to one independent resonator for each animal. In this work, multiple-animal and phased-array technologies were combined to further enhance the imaging efficiency and flexibility of small-animal MRI. Multiple arrays of receive coils (MARCs) were used for accelerated anatomical imaging of two mice in vivo. Reconstructed images from a sensitivity encoding (SENSE) algorithm demonstrate the feasibility of using MARCs to substantially improve the efficiency and reduce the cost of small-animal MRI.
MR Systems, Magnets, Gradients, Shims
Hall D Tuesday 13:30-15:30
Franco Bertora1, Elisa Molinari1, Andrea Viale1
1Italian Institute of Technology, Genova, Italy
Permanent magnet design for MRI has always followed lines that are forcedly different from those employed in designs using superconductors yet, from a basic point of view, the physical principles are the same and a block of permanently magnetized material can be considered as the epitome of ambient temperature superconductivity. When confronted with the problem of designing an open scanner for functional imaging the techniques employed in designing permanent magnets can advantageously be applied to superconducting windings, particularly now that emerging materials such as MgBr2 afford relative ease in the construction of cryogen-free devices.
1149. An Optimised Elliptical Magnet for Deep Surface NMR Imaging
Manola Ciarrocchi1, Angelo Galante1, Vincenzo Di Miccoli2, Marcello Alecci1, Antonello Sotgiu1
1University of L'Aquila, L'Aquila, Italy; 2Itel Telecomunicazioni, Ruvo di Puglia, Italy
Unilateral NMR allows positioning of the sample on the surface of a portable measuring device and this technique has been used for several MRS/MRI applications. However, the open magnet geometry produces a magnetic field with high inhomogeneity, giving rise to reduced relaxation time T2*. MRI applications require a good field homogeneity and increased penetration depth. Unfortunately, most of the unilateral devices present in the literature do not satisfy both these requirements. Here we report the design of a novel elliptical unilateral magnet with improved field homogeneity and penetration along one direction.
1150. Shimming a 0.2 T Permanent Imaging Magnet with Small NdFeB Magnets
David Ian Hoult1, Qunli Deng2, Boguslaw Tomanek2
1National Research Council Institute for Biodiagnostics, Winnipeg, Canada; 2National Research Council Institute for Biodiagnostics, Calgary, Canada
Field shimming (20 ppm over a 30 cm diameter sphere) with small NdFeB magnets on the pole faces of an inexpensive, 0.2 T, permanent imaging magnet is described. Problems overcome included: 950 ppm starting inhomogeneity; drift during field plotting; magnetic moment inconsistency; unknown, induced local pole-face magnetisation; change of magnetisation with magnet inversion, and insufficient convergence of spherical harmonic amplitude with order for magnets close to pole face centres. Keys to success were the modelling of pole face magnetisation by a perpendicular line of magnets and the use of constrained linear programming to maximise the number of shimming magnets.
1151. Characterization of MRI Properties of Human Body Tissues at MicroTesla Magnetic Fields
Byeong-Ho Eom1, Mark Steven Cohen2, Inseob Hahn1, Konstantin I. Penanen1
1California Institute of Technology, Pasadena, California , USA; 2University of California, Los Angeles, Los Angeles, California , USA
Using an imager operating at 170 microTesla we studied the in vivo relaxation characteristics of human tissue and discuss the contrast range and instrument optimizations necessary to provide quality low field imaging
Christoph Barmet1, Nicola De Zanche2, Bertram Wilm2, Klaas P. Pruessmann2
1University and ETH Zürich, Zürich, Switzerland; 2University and ETH Zürich, Switzerland
Magnetic field monitoring with NMR probes is a promising approach but suffers from practical limitations when performed with receive-only probes, relying on external RF excitation. This contribution describes the implementation of a transmit/receive monitoring system, relying on autonomous probeheads equipped with individual RF excitation and RF shielding against MR contamination from the actual imaging experiment. The newly designed probes are found to offer sufficient sensitivity and RF shielding, enabling straightforward monitoring of MRI procedures, as exemplified by initial phantom scans.
1153. Residual Magnetism in MR Suites After Field Rampdown of Superconducting Magnets
Steffen Sammet1, Francisco Aguila1, Regina Maria Koch1, Michael Vincent Knopp1
1The Ohio State University, Columbus, USA
The rampdown of two superconducting clinical magnets, one at 8T and one at 0.7T, was used to evaluate residual magnetization within the MRI suite environments prior, during and after field-rampdown. A controlled rampdown of even an ultrahigh field MR system does not lead to retained magnetic contamination, while forced quenched rampdown of a mid-field system revealed temporary negative remanence.
Silke Maria Lechner1, 2, Adam B. Kerr3, Pekka T. Sipilä1, 2, Rolf F. Schulte1, Dirk Lange1, Florian Wiesinger1
1GE Global Research, Munich, Germany; 2Technical University Munich, Munich, Germany; 3Stanford University, Stanford, USA
Magnetic Field Monitoring (MFM) has been used as k-space trajectory measurement technique, whereupon the calibrated information is used in image reconstruction and result in reduced imaging artifacts like blurring and geometric distortions. This work presents a comparison of MFM with an alternative measurement method commonly used in MR. It is shown that higher image quality and less blurring artifacts are achieved using MFM. The comparison is evaluated in multi shot spiral out and echo-planar imaging sequences and demonstrated to substantially reduce artifacts in both, spiral and EPI. Image entropy is used as a quantitative quality metric.
1155. Gradient Linear System Modeling Using Gradient Characterization
Joseph Y. Cheng1, Borjan Aleksandar Gagoski1, Divya S. Bolar1, Christina Triantafyllou2, 3, Michael Hamm4, Gunnar Krueger5, Elfar Adalsteinsson1
1MIT, Cambridge, Massachusetts, USA; 2Athinoula A. Martinos Imaging Center, McGovern Institute for Brain Research, MIT, Cambridge, Massachusetts, USA; 3Athinoula A. Martinos Center for Biomedical Imaging, Department of Radiology, MGH, Charlestown, Massachusetts, USA; 4Siemens Medical Solutions, Charlestown, Massachusetts, USA; 5Siemens Medical Solutions, Lausanne, Switzerland
The gradient system is modeled here as a linear time-invariant system, H(f), through a frequency-domain analysis. The model is obtained using a recent gradient characterization technique that combines two previous methods to optimize flexibility, speed, and accuracy. The obtained H(f) generalizes the gradient system to allow for a fast and accurate distortion prediction. As a result, the model will contribute to a robust correction method.
1156. Compensation of Eddy Current by an R-L-C Circuit Model of the Gradient System
Sang Heom Cho1, Pan Ki Kim1, Jong Woo Lim2, Su Yeol Jeon1, Chang Beom Ahn 3
1Kwangwoon University, Seoul, Republic of Korea, 2ISOL Technology
The k-space trajectory is important in the design of spiral pulse sequence as well as in the reconstruction of the image. The real trajectory is, however, usually deviated from the theoretical trajectory obtained from the gradient waveforms due to the eddy currents and non-ideal performances of the gradient systems such as finite bandwidth and slew rate. Such deviations inevitably result in distortions in the reconstructed image. In this paper, we derived an R-L-C circuit model to estimate the real k-space trajectory, by which a significant improvement of reconstruction was achieved at 3 Tesla MRI system.
1157. Reduction of MRI Scanner Acoustic Noise Using a Micro-Perforated Panel Absorber
Michael Li1, Chris K. Mechefske1
1Queen's University, Kingston, Canada
The trend toward higher field strength is worsening the noise problem in MRI. A Boundary Element Method simulation showed that a micro-perforated panel acoustic absorber can significantly reduce MRI noise in the scanner bore. The impedance function of the absorber was used to predict the sound attenuation effect. However, the absorption coefficient functions calculated according to Maa’s theory do not reflect the absorbing effect of a micro-perforated panel absorber in a cylindrical duct such as an MRI scanner bore because the expression of impedance of air gap is based on the assumption of plane waves
1158. Vibration Induced Eddy Current and Its Effect on Image Quality for MRI System
Longzhi Jiang1, Tim Havens2, William Einziger2
1GE Healthcare, Florence , USA; 2GE Healthcare, Florence, USA
During normal operation, MRI systems experience vibration from different external sources, e.g. refrigeration unit, environmental ground vibration and gradient coil pulse. Moving conductors within the MRI system will generate eddy currents, disturbing the homogenous magnetic field and affecting image quality (so called “ghosting”). In this study, vibration induced magnetic field fluctuation was investigated on a system with a single degree of freedom to illustrate the behavior. The methodology was then extended to a 3 dimensional MRI system by using FEM/BEM. Experimentally simulated results are presented, showing general agreement between experimental results and simulation for the 1.5T MRI system.
1159. Control of Gradient Coil Natural Frequency Using a Topology Optimization Technique
Sun Yong Kim1, Il Yong Kim1, Chris K. Mechefske1, Doo Ho Lee2
1Queen's University, Kingston, Canada; 2Dong-eui University, Busan, Republic of Korea
By controlling the natural frequency of the single-winging gradient coil, it is possible to reduce vibration resonant frequency amplitude. The overall noise and vibration of the structure can also be reduced. Viscoelastic damping material is widely used to suppress noise and vibration in industry due to its easy application and economic advantage. A full treatment of the damping material on the gradient coil does not give a maximum damping effect for the structure. The optimum location, thickness and shape of the damping sheets are being determined using topology optimization. This work aims to minimize the level of noise and vibration of the structure.
1160. Phantom Correction of Human Images for Spatial Scaling Errors
Jeff Gunter1, Matt Bernstein2, Bret J. Borowski2, Paula J. Britson1, Chadwick P. Ward2, Clifford R. Jack, Jr. 2
1Mayo Clinic and Foundation, Rochester, Minnesota, USA; 2Mayo Clinic, Rochester, Minnesota, USA
An investigation of the applicability of phantom-derived scanner scaling estimates close-in time human images is presented. Phantom correction reduces human image scaling variability within and across scanners.
1161. Frequency Spectrum of Partial Discharge Events
Derek A. Seeber1, Tony M. Linz2, Bruce D. Collick2, Anthony Mantone1
1GE Healthcare, Florence, South Carolina, USA; 2GE Healthcare, Waukesha, Wisconsin, USA
Each gradient coil, inner and outer, consists of three layers corresponding to the X, Y, and Z-axes of the gradient coil. The inner and outer gradient coils each have two radial spaces between adjacent axes. The radial spacing between adjacent axes is required to reduce the occurrence of “white pixels” and voltage breakdown during an MRI scan. If the “white pixel” performance of the gradient coil can be increased, the radial space savings can be incorporated into an increased patient bore, reduced magnet expenditure, or increased gradient slew rates with higher applied gradient voltages.
1162. Modal Sound Radiation from Finite Cylindrical Shells
Tian Ran Lin1, Peter O'Shea1, Chris K. Mechefske2
1Queensland University of Technology, Brisbane, Australia; 2Queen's University, Kingston, Canada
Modal radiation efficiencies of two fundamental circumferential modes of several different length finite cylindrical shells are studied in this paper. Modal radiation efficiencies of the finite cylinders are found to asymptotically approach those of the corresponding infinite cylindrical shell when the length of the cylinder divided by the circumference modal index (n>1) is more than twice that of the acoustic wavelength. There are two radiation peaks in the modal radiation efficiency attributed to the ring and critical frequencies of the finite cylinder.
1163. Initial Realisation of a Multichannel, Non-Linear PatLoc Gradient Coil
Anna Masako Welz1, Maxim Zaitsev1, Heinrich Lehr2, Gerrit Schultz1, Zhenyu Liu3, Feng Jia3, Hans Post2, Jan Korvink3, Jürgen Hennig1
1University Hospital Freiburg, Freiburg, Germany; 2Bruker BioSpin MRI GmbH, Ettlingen, Germany; 3University Freiburg, Freiburg, Germany
This abstracts presents the initial realisation of image encoding with a multichannel, non-linear PatLoc (parallel imaging in localized gradients) gradient coil and first reconstructed images. Imaging in non-linear, non-bijective magnetic fields has the advantage of faster imaging, or higher resolution especially at the periphery. To conserve some image reconstruction properties, in plane orthogonal gradients were chosen, in radial and circumferential direction. A cylindrical concept was developed for later use as a head gradient set. The first demonstrator is designed for use in an animal system for easier access to the hardware.
1164. Parametrical Optimization of a PatLoc Gradient Coil
zhenyu Liu1, Feng Jia1, Maxim Zaitsev2, Anna Welz2, Gerrit Schultz2, Jan G. Korvink1, Juergen Hennig2
1Dept. of Microsystems Engineering, Laboratory for Simulation, University Freiburg, Freiburg, Germany; 2Dept. of Diagnostic Radiology, Medical Physics, University Hospital Freiburg, Freiburg, Germany
A simple iterative optimization method, instead of the commonly used target field method, is presented to design PatLoc gradient coil. The optimization procedure includes the calculation of the magnetic value using the Biot-Savart method, the determination of the optimal number of conductor should be used, and the corresponding space positions. In order to use the continuous optimization method to optimize the discrete conductors, the continuation method is used to transform the integral design variable to bounded continuous design variable. The optimal solution still keeps the discrete conductor distribution so that the optimized gradient coil can be fabricated without further post-processing.
1165. Shoulder-Slotted Insertable Gradient and Shim Coil Se
Michael Poole1, Dan Green2, Richard Bowtell1
1University of Nottingham, Nottigham, UK; 2Magnex Scientific Ltd., Yarnton, UK
Insert gradient and shim coils designed specifically for head imaging can be made smaller and therefore of lower inductance and resistance as well as capable of producing more intense magnetic fields. Here, an Inverse Boundary Element Method was used to design a cylindrical, shoulder-slotted 3-axis head gradient coil set, also containing 0th and 2nd order shim coils, The shim coils were designed to have low-inductance so as to allow the fast current switching required by dynamic shim updating. The gradient and shim coil set has been constructed and tested at 3T in imaging experiments.
1166. Dynamic Shimming at 7 Tesla
Saikat Sengupta1, Yansong Zhao2, David Foxall2, Piotr Starewicz3, Adam Anderson1, John Gore1, Malcom Avison1
1Vanderbilt University Institute of Imaging Science, Nashville, USA; 2Philips Medical Systems, Inc, USA; 3Resonance Research, Inc, Billerica, USA
We have implemented dynamic shimming on Philips Achieva human 7Tesla system A Real Time Shim System, RTS, (Resonance Research Inc, MA, USA), was used for higher order shim control. Shim corrections were calculated from preacquired fieldmaps. The 1st order corrections were loaded slice wise through the gradient controls and the 2nd order shims were loaded from the RTS system. The static field standard deviations across the brain in all slices were significantly lower with dynamic shimming than with global shimming. This improvement in B0 homogeneity was reflected in reduced signal drop-out and lower image distortion.
1167. Power Versus Inductance: Finite Length Shim Coil Design for High-Field MRI
Parisa Jamali1, Blaine A. Chronik1
1University of Western Ontario, London, Canada
We designed high power, high order shim sets with constrained length using two methods: the minimum inductance method and the minimum power method. While the minimum inductance method gave higher inductive merit and the minimum power method gave higher resistive merit, the percent difference in merits between the two methods was less than six percent. This indicates that we can use the minimum power method for the design of our shim axes.
1168. Local Uni-Planar Gradient Array Design Using Conformal Mapping and Simulated Annealing
Sung M. Moon1, D.N. Ghosh Roy1, K. Craig Goodrich1, J. Rock Hadley1, Dennis L. Parker1
1University of Utah, Salt Lake City, Utah, USA
Many imaging problems, such as breast lesion characterization, can require high spatial and temporal resolution. The improved gradient performance required to achieve high spatial and temporal resolution may be achieved by specially designed local gradient coils such as planar gradient inserts. We present simple and rapid method for design of flat gradient inserts to produce a high strength gradient field and a reasonably uniform imaging region. The planar gradient set is to be placed inside of the imaging bore of the magnet (within the body gradients). For cylindrical gradients, a one dimensional stream function (SF) is used to specify currents on the cylinder surface. For planar gradients, however, this becomes a more complicated two dimensional problem.
1169. Effect of Active Shielding on Zonal Shim Coils for a 31cm Bore 9.4T MR System
Dustin Wesley Haw1, Blaine Alexander Chronik1
1University of Western Ontario, London, Canada
MRI and in-vivo spectroscopy in preclinical applications at very high field (9.4T and above) require improved shimming capabilities. Our goal is to develop and optimize high strength, dynamic shim systems for our 31cm bore 9.4T MR system. It is clear that dynamic shim systems will need to be actively shielded; however, we are interested in investigating how shim performance of shielded shim coils changes as a function of shim order. In all cases the efficiency of unshielded shim coils is better than that of shielded shim coils, as expected. However, the difference decreases considerably with increasing shim order.
1170. MR Imaging Capability of a Field-Cycled MRI/PET Scanner
Kyle Michael Gilbert1, Timothy James Scholl1, Blaine Alexander Chronik1
1The University of Western Ontario, London, Canada
Field-cycled MRI employs two separate, actively controlled resistive magnets to polarize a sample and to provide the magnetic field environment under which data is acquired. A field-cycled MRI system was constructed with a central gap to allow for the inclusion of a PET ring. The SNR and magnetic field homogeneity of the field-cycled MRI system were sacrificed to create a dual-modality scanner. To compensate, a low-pass birdcage coil and a higher order shim set were built, allowing for the acquisition of quality MR images.
1171. From Static to Dynamic: Construction of a B0 Insert for Field-Cycled Contrast in a Clinical MR Scanner
Jamu K. Alford1, Timothy J. Scholl1, William B. Handler1, Brian K. Rutt2, Blaine A. Chronik1
1The University of Western Ontario, London, Canada; 2Robarts Research Institute, London, Canada
To allow field-cycled contrasts in superconducting MR scanners, we have built an actively shielded B0 insert. This insert produces significant magnetic field shifts in the imaging region of clinical MR systems without destabilizing their main field or producing eddy-currents. The insert coil is composed of a powerful primary coil responsible for the B0 field shift and a precisely designed, counter wound, outer shield designed to reduce the insert's fringe field. Design, construction and testing of this unique insert are discussed.
1172. Towards Simultaneous PET and Field-Cycled MRI: Active Shielding for PMT Detectors
Dustin Wesley Haw1, Blaine Alexander Chronik1
1University of Western Ontario, London, Canada
Combined PET and MR scanners are currently under development as a means to obtain specific functional data from PET, registered both temporally and spatially with high-resolution anatomical images from MRI. One approach is to use PET and field-cycled MRI (FCMRI), which uses two separate and dynamically controlled magnets for the polarization and readout phases of MRI, in an interleaved manner. Actively shielded polarizing magnets reduce the magnetic field over the region in which the PMTs would be placed, and allow for extended PET operation during the polarizing phase of FCMRI.
1173. Quantitative Signal and Phase Analysis of a Field-Cycled MRI Scanner
Kyle Michael Gilbert1, Timothy James Scholl1, Jamu Krishna Alford1, Blaine Alexander Chronik1
1The University of Western Ontario, London, Canada
Field-cycled MRI employs two actively controlled electromagnets to polarize a sample and to provide the magnetic field environment during data acquisition. Instabilities in the readout magnet can cause phase errors and image artifacts in the phase-encode direction. The stability of the readout magnet was quantitatively evaluated, along with the expected signal trends. Images were acquired with no visible phase-encode artifacts.
1174. Preliminary Results of CdTe Detector Capabilities Toward MRI-SPECT
Takao Goto1, Yuji Iwadate1, Yoshio Mito2, Tetsuji Tsukamoto1
1GE Yokogawa Medical Systems, Hino-shi, Japan; 2Acrorad co. ltd, Uruma-shi, Japan
Recently there have been new interests in MR Hybrid Systems such as MR-PET and MR-SPECT. CdTe detector has valuable potentials as a new radiation detector and does not need Photo Multiplier. We investigated the interference of MR and CdTe detector toward MR-SPECT realization measuring Enegry Spectra changes, imaging Co-57 point source and scanned MR phantom with Tungsten laminated collimator in both 1.5T and 3T magnetic field. As a result, it is demonstrated that there was no significant effect in working CdTe detector in both 1.5T and 3T and found collimator effect could be avoided to scan specified imaging plane.
1175. Design and Fabrication of a Magnetic Resonance Stage Microscope
Andrey V. Demyanenko1, Julian Michael Tyszka1
1California Institute of Technology, Pasadena, California , USA
A uniplanar geometry analogous to an inverted stage optical microscope is proposed for high resolution MR imaging which addresses many of the limitations of conventional volume gradient and RF coil designs, particularly restrictions on medium volume and physical access when imaging small organisms such as developing embryos or thin tissue explants. The stage microscope geometry provides unhindered access to the sample from above, encouraging the future integration of optical imaging equipment with MR microscopy.
1176. Multiple Mouse Imaging of 16 Live Mice
Jonathan Eric Bishop1, Shoshana Spring1, Jun Dazai1, Brige Paul Chugh1, Sharon Portnoy1, Steven Suddarth, G Ronald Morris, R Mark Henkelman1
1Hospital for Sick Children, Toronto, Canada
Preliminary results for fully parallel imaging of 16 live mice are presented. The technique scales well and provides a significant increase in throughput compared to single mouse imaging.
1177.
Development of a Compact Whole Hand MRI System for Diagnosis
of Rheumatoid Arthritis Using a 0.3 T
Permanent Magnet
Shinya Handa1, Katsumi Kose1, Tomoyuki Haishi2
1University of Tsukuba, Tsukuba, Japan; 2MRTechonology Inc., Tsukuba, Japan
A compact whole hand MRI system for diagnosis of rheumatoid arthritis (RA) was developed using a 0.3 T permanent magnet and a locally shielded RF probe. The total system was installed in a 1.5 m ´ 2.0 m open space. Because the T1 weighted 3D images showed 1.6 times SNR advantage over those obtained with a 0.21 T permanent magnet previously reported, we concluded that our new system can be used for evaluation RA with reduced examination time or higher spatial resolution in the same examination time.
Hall D Tuesday 13:30-15:30
Leslie Ying1, Jinhua Sheng1
1University of Wisconsin - Milwaukee, Milwaukee, Wisconsin, USA
JSENSE formulates parallel imaging as a nonlinear problem to estimate the coil sensitivities and the desired images simultaneously. The existing algorithm solves the problem by iterative alternating minimization, which requires quite a few self calibration data for an accurate initial sensitivity estimation. In this abstract, we propose to use variable projection method to solve the nonlinear optimization problem. This method requires very few self calibration data because it converges to an optimal solution regardless of the initial value. The proposed method has been tested on a set of simulation data and demonstrated promising results.
1260. Automatic Coil Selection for SENSE Imaging with Large Coil Arrays
Mariya Doneva1, Peter Börnert2
1University of Luebeck, Luebeck, Germany; 2Philips Research Europe, Hamburg, Germany
The use of a large number of coil elements for parallel imaging allows improved imaging performance and increased SNR but can lead to memory storage problems and increased reconstruction times. We present an efficient approach for coil selection, based on singular value decomposition (SVD), applicable to massively parallel SENSE imaging. The SVD-based coil selection can be used instead of manual coil selection in conventional scan planning and is especially useful in planning double oblique SENSE scans. It is also applicable to real-time or interventional imaging, where the selection could be performed locally enabling dynamical coil switch during image acquisition.
1261. A Comparison of Reconstruction Techniques for Non-Uniformly Sampled 3D Parallel Imaging
Zhikui Xiao1, 2, William Scott Hoge2, R. V. Mulkern, 23, Guangshu Hu1, Walid E. Kyriakos2, 3
1Tsinghua University, Beijing, People's Republic of China; 2Brigham and Women's Hospital, Boston, Massachusetts, USA; 3Children's Hospital, Boston, Massachusetts, USA
In this abstract, we explore a specific non-uniform sampling scheme for 3D imaging on a rectilinear grid. We show that high-quality images can be reconstructed by 2D-SPACE RIP and 2D-GRAPPA-Operator. To evaluate the proposed sampling method and reconstruction schemes, results from a phantom study and in-vivo 3D human data are shown. Overall, fewer artifacts can be seen in the 2D-SPACE RIP reconstructions.
1262. Accelerating Acquisition in Spiral Imaging
Ajit Devaraj1, Jim Pipe1
1St. Joseph's Hospital and Medical Center, Phoenix, Arizona , USA
Most present day 3D imaging techniques are based on non-cartesian trajectories to help traverse k-space rapidly. The ever increasing demand for reduced scan time makes non-cartesian parallel imaging necessary. However non-cartesian parallel imaging is computation intensive and algorithmically complex. PILS-like algorithms provide a good compromise between potential acceleration factors and the complexity introduced in reconstruction. Presented here is a PILS inspired algorithm for multi-coil reconstruction of spiral data acquired with a bird-cage coil geometry.
1263. Parallel Image Reconstruction Using a Single Signal and PSFT Anti-Alias Image Reconstruction
Satoshi Ito1, Yoshifumi Yamada1
1Utsunomiya University, Utsunomiya, Japan
Sensitivity encoding (SENSE) accelerate MR scan time by using multiple receiver coil in parallel scan time. Here, we propose a method to reconstruct under-sampled images using only a single set of signals. The signal obtained in the phase-scrambling Fourier Transform imaging (PSFT) can be transformed into the signal described by the Fresnel transform of the objects, in which alias-free images can be obtained by optionally scaling the object images to fit in the field-of-view (FOV). By applying a weighting function to the alias-free image corresponding to sensitivity of the receiver coil and then calculating a weighted PSFT signal from the weighted alias-free images, we can obtain two or more signals with different sensitivity distributions. The proposed method is demonstrated using 2-fold under-sampling with k-space trajectories.
1264. Non-Cartesian Parallel Reconstruction Using Null Operations (NC-PRUNO)
Jian Zhang1, Chunlei Liu1, Michael Moseley1
1Stanford University, Stanford, California , USA
A new k-space based parallel reconstruction method is proposed which is called Non-Cartesian Parallel Reconstruction Using Null Operations (NC-PRUNO). This method can be used for interleaved trajectories with auto-calibration such as variable density spirals. In NC-PRUNO, all missing interleaved k-space samples are synthesized from solving a linear equation, which is formulated by applying some local null operators on all k-space locations. We also demonstrate that it can be solved efficiently and accurately with a conjugate gradient method. In comparison with other k-space methods, NC-PRUNO tries to directly estimate the missing interleaves while strictly maintaining the consistency of the acquired samples.
1265. Reconstruction of Phase Images for GRAPPA Based Susceptibility Weighted Imaging (SWI)
Christian Ros1, Stephan Witoszynskyj1, Karl-Heinz Herrmann1, Jürgen R. Reichenbach1
1Friedrich-Schiller-University, Jena, Germany
Susceptibility Weighted Imaging (SWI) relies on both magnitude andphase information. Long acquisition times make SWI an interestingapplication for parallel imaging techniques, such as GRAPPA. However,the use of phased arrays requires schemes for combining images of theindividual channels. We have implemented a method that usessensitivity information from a separate measurement within the Siemens image-reconstruction framework. The results of this method were compared tothe standard Adaptive Combine reconstruction and to data obtained withthe body-coil. Images computed with our method were in good agreementwith the body-coil's data. Adaptive Combine did not only frequentlyfail to reconstruct phase images correctly, but also caused severeartifacts in magnitude images.
1266. Automatic Time Frames Subset Selection for Improved TGRAPPA Reconstruction
Roger Nana1, Xiaoping Hu1
1Georgia Institute of Technology / Emory University, Atlanta, USA
In TGRAPPA, several adjacent undersampled time frames are combined to compose the data used for deriving the weights for the GRAPPA reconstruction. In this work, a method that exploits the shift invariance property of Cartesian GRAPPA is introduced to automatically select the set of time frames for the calibration that results in an optimal TGRAPPA reconstruction. This data-driven approach was demonstrated with experimental data to lead to improved performance.
Yoon Chung Kim1, Jeffrey A. Fessler1, Douglas C. Noll1
1University of Michigan, Ann Arbor, Michigan, USA
In parallel imaging, sensitivity map is widely used as calibration data which plays a crucial role during the unaliasing process. Sensitivity map estimation process usually includes spatial/temporal smoothing in order to improve robustness to noise or motion artifacts. However, the effect of smoothing has not been thoroughly investigated to date. In our study, we investigate the effects of both spatial and temporal smoothing of sensitivity maps on motion corrupted fMRI data. We also propose a novel self-calibrated sensitivity map estimation technique that controls noise and smoothness via regularization.
1268. A Self-Calibration Technique for Suppression of Radial Undersampling Artifacts in Parallel Imaging
Yu Li1, Feng Huang1, Randy Duensing1
1Invivo Diagnostic Imaging, Gainesville, USA
In this study, a self-calibration technique was developed to suppress the radial undersampling artifacts for parallel imaging. This technique takes advantage of an intrinsic property of radial sampling and does not need any extra calibration data. It was demonstrated that the streak artifacts can be efficiently suppressed using this method in 2D and 3D (VIPR) radial imaging.
1269. Phase Constrained Parallel Imaging for Improved Fat Suppression in Multi-Echo LC SSFP
Youngkyoo Jung1, Alexey A. Samsonov1, Richard Kijowski1, Jessica L. Klaers1, Walter F. Block1
1University of Wisconsin-Madison, Madison, Wisconsin, USA
The Linear Combination SSFP (LC-SSFP) method that combines two steady-states with different passbands provides fat/water separation. However, unwanted bright fat signal is not fully suppressed. With proper phase constraints in voxels with partial voluming, the residual unwanted fat signal in water volume can be effectively removed. In this work, we demonstrate an improved SMART CG algorithm to provide parallel imaging, phase coherence, and markedly improved fat/water separation in LC SSFP images acquired with dual half-echo 3D PR trajectories. Our result demonstrates substantial image quality improvement in musculoskeletal images.
1270. A Sparse TSENSE Approach for Improved Dynamic Parallel MRI
Martin Blaimer1, Felix A. Breuer1, Peter M. Jakob1, Peter Kellman2, Mark A. Griswold3
1Research Center Magnetic Resonance Bavaria (MRB), Würzburg, Germany; 2National Institutes of Health, National Heart, Lung and Blood Institute, Bethesda, USA; 3University Hospitals of Cleveland and Case Western Reserve University, Cleveland, USA
In dynamic parallel MRI improved reconstruction quality can be achieved by taking into account that signal changes occur in localized regions only. Here, a frame-work for improved TSENSE reconstructions is presented. TSENSE is based on a time-interleaved acquisition scheme and does not require a separate pre-scan for coil sensitivity estimation. Thus, robust reconstructions at high frame rates can be obtained. It is shown, that the noise enhancement due to ill-conditioning of the inverse problem is significantly reduced by considering only dynamic regions for the reconstruction process. A simple algorithm is presented for determining the dynamic regions and reducing the ill-conditioning.
1271. Highly Accelerated 2D GRAPPA by Randomized K-Space Sampling
Clifton R. Haider1, Stephen J. Riederer1
1Mayo Clinic, Rochester, Minnesota, USA
With typical Cartesian acquisition the maximum acceleration along a single direction is more often limited by the inability to unalias the image than by the noise amplification. Because acceleration techniques are generally more robust when performed along two vs. one direction, the initial implementation of random sampling GRAPPA studied uses 2D acceleration. Further, a Gaussian distribution is applied to differentially weight the sampling of the lower spatial frequencies to improve image quality. We show experimental studies done in phantoms using eight coils with net acceleration factors, including the autocalibration region, of 4 and 8.
Satoshi Ito1, Yoshifumi Yamada1
1Utsunomiya University, Utsunomiya, Japan
A novel image reconstruction technique is proposed in which parallel image reconstruction is performed based on the SENSE algorithm using only a single set of signals. The signal obtained in the phase-scrambling Fourier transform imaging technique can be transformed to the signal described by the Fresnel transform of the objects, which is known as the equation of diffracted wave-front equation of the object in acoustics or optics. The application of a weighting function to the PSFT signals has a similar effect as the application of a sensitivity function to the object function itself. Therefore, we can obtain two or more folded images from a single set of signals, and image reconstruction based on the SENSE algorithm is possible using a series of folded images given different weighting functions.
1273. Variable Density Sampling in Radial K-T GRAPPA
Jingsi Xie1, Feng Huang2, Kui Ying3, Debiao Li1
1Northwestern University, Chicago, Illinois, USA; 2Invivo Corporation, Gainesville, USA; 3Tsinghua University, Beijing, People's Republic of China
Radial GRAPPA and radial k-t GRAPPA have been proposed in radial imaging. But they need extrally acquired or calculated full k-space data, which require more reconstruction time. In this abstract, we show that variable density sampling is applicable in radial k-t GRAPPA, which avoids the extra acquisition or calculation of calibration signal.
1274.
Channel Compression for BLADE
Alto Stemmer1, Vladimir Jellus1, Stephan Kannengiesser1, Berthold Kiefer1
1Siemens Medical Solutions, Erlangen, Germany
Novel multi-channel coils with 32 or more elements allow shorter scan times due to improved signal to noise ratios and higher parallel imaging performance. Increased reconstruction times for advanced imaging techniques such as BLADE can, however, be a problem in routine clinical imaging. In this work, we show that the reconstruction times can be reduced by introducing channel compression techniques at various points in the BLADE reconstruction pipeline.
1275. Phase-Constrained Reconstruction of GRAPPA for Accelerated MR Acquisitions
Mohammad Sabati1, 2, Haidong Peng1, 2, Richard Frayne1, 2
1University of Calgary, Calgary, Canada; 2Calgary Health Region, Calgary, Canada
Scan time reduction is important in clinical MR imaging. Partial Fourier acquisitions rely on the conjugate symmetry of Hermitian data and allow for shorter scan times with fewer phase-encoding steps. A further scan time reduction has been achieved by combining partial k-space acquisitions with parallel imaging. In this study, we propose a combined GRAPPA+POCS (GRAPOCS) technique that allows simple and efficient inclusion of phase- and data-consistency constraints in reconstruction to improve image quality and to achieve higher acceleration factors. We evaluate the performance of this method using phantom and human data.
1276. GRAPPA-POCSENSE
Eugene G. Kholmovski1
1University of Utah, Salt Lake City, USA
GRAPPA is a widely used technique for parallel MRI. The technique usually has sub-optimal SNR and residual aliasing in comparison with SENSE. One of the main causes for the GRAPPA sub-optimality is an inconsistency between individual coil images which reconstructed independently one from the others. POCSENSE is an iterative technique for parallel MRI that enforces consistency between individual coil images and allows incorporation of various constraints to improve resulting image quality. GRAPPA and POCSENSE can be readily combined to utilize strong points of both methods for improved parallel MRI reconstruction.
1277. An Approach to Coil Calibration Based on Prior Training Data
Francesco Padormo1, Rita G. Nunes1, David Atkinson2, Philip G. Batchelor, Jo V. Hajnal1, David J. Larkman1
1Hammersmith Hospital, Imperial College London, London, UK; 2University College London, UK
Calibration of coil sensitivities is an intrinsic part of the MR exam. Calibration is used for parallel imaging or to correct signal modulations due to local coils. As coils become smaller, higher resolution images are needed for calibration and scan times increase. Parallel transmit further increases the time taken to calibrate as both the transmit (B1) and receive fields are required. This work explores the variation in receive coil sensitivities across a population and using this to construct a model of coil sensitivity which could be used to reconstruct unknown coil sensitivities, reducing or removing the need for individual calibration.
1278. Improved Spatial Homogeneity and Sensitivity Estimation for Multi-Coil Image Reconstruction
Tolga Çukur1, Dwight Georger Nishimura1
1Stanford University, Stanford, California , USA
Receiver arrays composed of surface coils generally yield non-uniform coverage over the spatial extent of the array. If the coil sensitivities are exactly known, then this non-uniformity can be corrected for in the reconstruction step. However, accurate sensitivity information is usually not available. For most applications, either a sum-of-squares (SOS) reconstruction is performed or an SOS normalization is used to estimate the sensitivities. Nevertheless, the SOS combination usually yields intensity modulations. A pth-norm combination that achieves a flatter profile can instead be employed to improve the sensitivity estimates. Both multi-coil image combination and self-calibrated parallel imaging applications are shown to benefit from the proposed method.
Anja C.S. Brau1, Philip J. Beatty1, Charles A. McKenzie2, Huanzhou Yu1, Ann Shimakawa1, Scott B. Reeder3, Jean H. Brittain4
1GE Healthcare, Menlo Park, California , USA; 2University of Western Ontario, London, Canada; 3University of Wisconsin, Madison, Wisconsin, USA; 4GE Healthcare, Madison, Wisconsin, USA
This work examines the impact of parallel imaging reconstruction on image phase to identify possible sources of phase perturbation. The phase effects of SENSE and ARC reconstruction are compared in the context of multi-point phase-sensitive imaging. It is shown that when the unaliasing step is separated from the coil combination step, for example in coil-by-coil reconstructions, original image phase is unperturbed. This finding has implications for the design of parallel imaging calibration strategies for phase-sensitive imaging.
1280.
Phase-Preserving Multi-Coil Combination with Improved
Intensity Modulation
Rexford David Newbould1, Stefan Skare1, Roland Bammer1
1Stanford University, Stanford, California , USA
Each coil image in a phased array contains a different phase offset, due to the coil's geometry, wiring, receiver delays, and electrical properties under varied loads. Coil images are usually combined using the magnitude sum-of-squares (SoS) operation, due to its ease and no a priori coil knowledge required. However, the SoS discards phase information, used in a number MR measurements, and exacerbates image intensity variation from the coil sensitivities which can be problematic in arrays with large numbers of small coils. A straightforward robust method for estimating coil phase offsets is presented, which can handle coils without direct signal overlap.
1281. Self-Calibrating Gridding for 3D Radial Trajectories Using GRAPPA Operator Gridding (GROG)Nicole Seiberlich1, Philipp Ehses1, Sonia Nielles-Vallespin2, Felix A. Breuer3, Martin Blaimer3, Peter M. Jakob1, 3, Mark A. Griswold4
1University of Wuerzburg, Wuerzburg, Germany; 2Siemens AG Medical Solutions, Erlangen, Germany; 3Research Center Magnetic Resonance Bavaria (MRB), Wuerzburg, Germany; 4University Hospitals of Cleveland, Cleveland, Ohio, USA
GRAPPA Operator Gridding (GROG) has been recently demonstrated as an alternative to convolution gridding. This method has a number of advantages over other gridding methods, but because it employs parallel imaging concepts to shift k-space points, a calibration dataset to determine the GROG weights is required. In two dimensional radial and spiral imaging, these weight sets can be determined by using the non-Cartesian datapoints themselves; this abstract demonstrates that an extension of this method can be used to determine GROG weights for 3D radial data. To this end, examples of an UTE dataset gridded with 3D SC-GROG are depicted.
1282. Description of Noise and Signal Probability Density Functions in SENSE Reconstructed Images
Alejandro Ribes1, I-Yun Chen1, Ching-Po Lin1
1National Yang-Ming University, Taipei, Taiwan
We describe the noise and signal PDF (Probability Density Function) for images reconstructed by SENSE. We formulate a hypothesis about the shape of the PDF using the Central Limit Theorem. For checking this hypothesis we experimentally approximate the distributions by histograms and fit different PDFs to them. The results indicate that Ricean PDFs appears in signal areas of the image and Rayleigh PDFs in image areas presenting no signal. These results are useful for the future development of methods to estimate Signal-to-Noise Ratio in SENSE reconstructed images and also for image post-processing algorithms.
1283. Sensitivity Encoding of Chemical Shifts
Justin P. Haldar1, Diego Hernando1, Dimitrios C. Karampinos1, Bradley P. Sutton1, John G. Georgiadis1, Zhi-Pei Liang1
1University of Illinois at Urbana-Champaign, Urbana, Illinois, USA
Conventional spectroscopic imaging experiments acquire multiple temporal encodings to enable the separation of different resonance frequencies. In this work, we explore a new kind of spectroscopic imaging that requires only a single temporal encoding, relying instead on the sensitivity encoding provided by an array of receiver coils. This provides a single-shot mechanism for chemical shift artifact correction and spectroscopic signal separation, although this comes at the expense of significant noise sensitivity.
Robin Martin Heidemann1, Fabrizio Fasano2, Markus Vogler3, 4, Christoph Leuze1, Josef Pfeuffer4, Robert Turner1
1Max-Planck Institute for Human Cognitive and Brain Sciences, Leipzig, Germany; 2Fondazione Santa Lucia, Rome, Italy; 3University of Applied Sciences, Hof, Germany; 4Siemens Medical Solutions, Erlangen, Germany
The use of parallel imaging can reduce blurring due to T2* relaxation and distortions due to off-resonance effects significantly. This can also be realized with a zoomed approach, eg using outer-volume suppression (OVS). However, both methods have their limitations. Due to imperfections in the reconstruction, parallel imaging can be affected by residual foldover artifacts. On the other hand, the zoomed images can be affected by remaining signal from imperfect OVS. The combination of both methods, in this work zoomed EPI with OVS and GRAPPA, is capable to deal with both problems at once and leads to increased image quality.
1285. G-Factor Maps of Conjugate Gradient SENSE Reconstruction
Bo Liu1, Emad Abdelsalam2, Jinhua Sheng1, Lei Ying1
1University of Wisconsin-Milwaukee, Milwaukee, Wisconsin, USA; 2GE Healthcare, Waukesha, Wisconsin, USA
In parallel imaging with Cartesian sampling, the spatially varying g-factor represents the loss in signal to noise ratio (SNR) due to ill-conditioning of the matrix inverse in SENSE reconstruction. In this abstract, we propose a method to calculate the spatially varying g-factor map for conjugate gradient (CG) SENSE reconstruction with arbitrary trajectories. The method allows us to analyze how different trajectories and number of iterations in CG affect the SNR in a spatially dependent way. Our experiments show that the g-factor maps of CG SENSE increase with the number of iterations.
1286. Characterization of Artifacts and Noise Enhancement Introduced by GRAPPA Reconstructions
Jonathan Rizzo Polimeni1, Graham C. Wiggins1, Lawrence L. Wald1
1MGH, Harvard Medical School, Charlestown, Massachusetts, USA
Two main sources of error---noise enhancement and image artifact---are introduced during accelerated parallel imaging. However few methods exist for characterizing the image reconstruction performance outside of the SENSE g-factor. Here we consider the noise enhancement and image artifact introduced by GRAPPA. Image artifact is quantified by validating the consistency of the GRAPPA kernel on fully-sampled data, and demonstrates an inability of the kernel to reconstruct high-frequency image features. Noise enhancement is measured through comparing image SNR between reconstructions from accelerated and unaccelerated acquisitions, and the spatial distribution of noise enhancement for GRAPPA is shown to differ substantially from the SENSE g-factor.
1287. Tikhonov Regularization: Effects on the Detection of Activations in SENSE Functional MRI
Alejandro Ribes1, Cyril Poupon2, Ching-Po Lin1
1National Yang-Ming University, Taipei, Taiwan; 2Commissariat a l'Energie Atomique, Saclay, France
We apply Tikhonov regularization to the SENSE reconstruction of fMRI series. We optimise the Tikhonov regularization parameter for the detection of such activations. Results on fMRI series at reduction factor 2 are shown.
1288. A Regularization with Prior Information Technique for GRAPPA
Feng Huang1, Yu Li1, George Randy Duensing1
1Invivo Corporation, Gainesville, Florida, USA
In this work, a regularized GRAPPA [2] with prior information technique is introduced. The regularization parameters are channel-wise, and are automatically determined by fitting the auto calibration signal (ACS) lines. Experimental results demonstrate that the proposed method can dramatically improve image quality, even if there is significant difference between the prior information and the target image. Moreover, the usage of low frequency prior information will not reduce the spatial resolution of the target image. The reconstruction time of conventional GRAPPA and the proposed regularized GRAPPA are almost idential.
1289. Local Mutual Information Guidied Denoising for Self-Calibrated PPI
Weihong Guo1, Feng Huang2
1University of Alabama, Tuscaloosa, Alabama, USA; 2Invivo Corporation, Gainesville, Florida, USA
Local mutual information (LMI) detects the similarity of two images. It is proposed in this work that the LMI between the image reconstructed by fully acquired central k-space data and the image by GRAPPA is used to detect the noise distribution and location of the edges, and then guide the adaptive noise suppression. Experimental results show that the proposed method significantly improved SNR without reducing the high frequency information.
1290. Temporal and Noise Behavior of PEAK-GRAPPA
Simon Bauer1, Matthias Honal1, Bernd Jung1, Jürgen Hennig1, Michael Markl1
1University Hospital Freiburg, Freiburg, Germany
PEAK-GRAPPA using a single uniform k-t GRAPPA is a promising tool for accelerated dynamic MR imaging. The purpose of this study was a detailed evaluation of the performance of PEAK-GRAPPA and the dependence of blurring and image quality on object dynamics. Results indicate that image quality, blurring, and noise behavior are considerably improved by PEAK-GRAPPA compared to standard methods. Noise performance demonstrated a clear dependency on object motion but was still superior to standard GRAPPA even at high object velocities.
1291. GRAPPA Navigators: Motion Correction with Parallel Imaging
Paden Roder1, Jacob Willig-Onwuachi1
1Grinnell College, Grinnell, USA
The SMASH navigator method for motion correction is adapted for use with GRAPPA. Simulations comparing the SMASH and GRAPPA versions are presented and discussed.
1292. Automated GRAPPA Kernel Selection Using Akaike Information Criterion
Keith Heberlein1, Roger Nana2, Stephen LaConte3, Xiaoping Hu2
1Siemens AG, Medical Solutions, Erlangen, Germany; 2Emory/GA Tech Dept. of Biomedical Engineering, Atlanta, Georgia, USA; 3Baylor College of Medicine, Department of Neuroscience, Houston, Texas, USA
GRAPPA reconstructions from parallel receivers rely on local mutual information between k-space neighbors acquired across multiple channels. Typically, the criteria for inclusion in the reconstruction kernel are based on ad hoc or empirical considerations. This work shows that the GRAPPA kernel selection can be framed as a model selection problem and that the well known Akaike Information Criterion (AIC) provides an automated and practical choice of reconstruction kernel.
1293. The GRAPPA Coefficients Estimation Using Weighted Least Squares Method
Eugene G. Kholmovski1, Sathya Vijayakumar1
1University of Utah, Salt Lake City, USA
GRAPPA is a widely used technique for parallel MRI. Essential component of this technique is an estimation of reconstruction coefficients from auto-calibrating data. Typically, least squares (LS) method is used to solve this over-determined problem. Weighted least squares (WLS) method is valid alternative to LS. The conventional WLS is formulated with weights proportional to signal SNR. The reason why this WLS formulation is not applicable for the GRAPPA coefficients estimation and the WLS modification to address the GRAPPA problem has been studied.
1294. Iterative GRAPPA (IGRAPPA) for Dynamic Parallel Imaging
Roger Nana1, Tiejun Zhao2, Keith Heberlein2, Sven Zuehlsdorff2, Renate Jerecic2, Xiaoping Hu1
1Georgia Institute of Technology / Emory University, Atlanta, USA; 2Siemens Medical Solutions, Malvern, USA
iGRAPPA is a recently introduced method in which other acquired lines in addition to the calibration lines are iteratively used to achieve an improved interpolation kernel for GRAPPA reconstruction. In this work, iGRAPPA is extended to real time non-gated non-breath-hold cardiac imaging in which the auto-calibrating lines for GRAPPA reconstruction are acquired only during the first time frame. The weights determined from the first time frame are iteratively updated based on data acquired at other frames in order to track the relative change in coil sensitivities. The effectiveness of this approach is demonstrated with in vivo data.
Philip M. Robson1, Aaron K. Grant1, Ananth J. Madhuranthakam2, Riccardo Lattanzi1, 3, Daniel K. Sodickson4, Charles A. McKenzie5
1Beth Israel Deaconess Medical Center, Boston, Massachusetts, USA; 2GE Healthcare, Boston, Massachusetts, USA; 3Harvard-MIT Division of Health Sciences and Technology, Boston, Massachusetts, USA; 4New York University School of Medicine, New York, New York, USA; 5University of Western Ontario, London, Canada
We demonstrate a general and robust method for the quantification of signal-to-noise-ratio (SNR) and g-factor for parallel MRI. The pseudo multiple replica SNR measurement described here allows image noise to be estimated on a pixel-by-pixel basis producing maps of SNR and g-factor that correctly represent the spatially-variant noise amplification of parallel imaging reconstruction techniques. This approach is universally applicable to any linear image reconstruction technique. We demonstrate SNR and g-factor maps for reconstructions from the same undersampled k-space data via the image-based generalized SENSE and the k-space-based GRAPPA techniques. The generality of this noise analysis method enables rigorous and quantitative comparisons between parallel imaging strategies, k-space trajectories, and image processing routines. We demonstrate SNR and g-factor maps in vivo which would have been impossible to obtain with either direct calculations or conventional multiple image replicas.
1296. Rapid Partially Parallel Reconstruction Using a Single Synthetic Target Coil
Weitian Chen1, Peng Hu1, Craig H. Meyer1
1University of Virginia, Charlottesville, Virginia , USA
In GRAPPA and BOSCO, convolution operations are performed on reduced FOV k-space data from each coil and then summed to obtain a full FOV k-space data set for an individual coil. The process is repeated for each coil and the images are then combined. We propose a novel parallel imaging method that formulates the convolution process in the image domain but with dramatically reduced computational cost, particularly when a large number of coils are used. We remove the outer loop over coils by generating one unaliased data set corresponding to a synthetic “coil”. The method is demonstrated using spiral acquisitions but is applicable to more general k-space sampling methods.
1297. Analytical Computation of G-Factor Maps for Autocalibrated Parallel Imaging
Philip James Beatty1, Anja C. S. Brau1
1GE Healthcare, Menlo Park, California , USA
This work describes how g-factor maps can be analytically and efficiently computed directly from the unaliasing coefficients used in autocalibrated parallel imaging. This direct approach allows the proposed method to be used with variable density and non-Cartesian acquisitions. The usefulness of g-factor maps as a tool for reconstruction parameter selection is examined in two important areas: selection of the coil combination method and selection of the reconstruction kernel size.
1298. Regularized Non-Cartesian SENSE Using a Multiscale Wavelet Model
Bo Liu1, Emad Abdelsalam2, Leslie Ying1
1University of Wisconsin - Milwaukee, Milwaukee, Wisconsin, USA; 2GE Healthcare, waukesha, Wisconsin, USA
The iterative conjugate gradient method (CG-SENSE) has been widely used for non-Cartesian SENSE. However, when the reduction factor is large, the ill-conditioning problem prevents the CG-SENSE from converging to the optimal reconstruction. To address this problem, we propose a regularization technique where the regularization term is a multiscale wavelet transform. The method has the advantage that the wavelet coefficients at different wavelet scales can be weighted by different regularization parameters such that both the smooth regions and sharp edges of images can be represented accurately. The experimental results demonstrate that the proposed method improves the convergence behavior and reconstruction quality.
1299. Improved Noise Performance Using Regionally Optimized Reconstruction for Partially Parallel Imaging
Yu Li1, Feng Huang1
1Invivo Diagnostic Imaging, Gainesville, USA
In this study, a regionally optimized reconstruction method for partially parallel imaging is introduced. Compared with the conventional SENSE, this method can reduce the ill-conditioning problems and improve the noise performance in the regions where g-factors are high. Compared with GRAPPA, this method minimizes the least-square error in reconstruction regionally instead of globally and hence gives better image quality in the regions where g-factors are low.
1300. Simultaneous Calibration Scheme for Data-Driven Parallel Imaging Reconstruction
Anja C.S. Brau1, Ann Shimakawa1, Philip J. Beatty1
1GE Healthcare, Menlo Park, California , USA
This work presents a flexible calibration strategy for data-driven parallel imaging whereby multiple datasets with potentially different magnetization can be used simultaneously to train the reconstruction. Such strategies could offer improved determination of reconstruction weights when calibration data from a single dataset is limited by motion, pulse sequence, or scan time. Technical feasibility is demonstrated in volunteer studies in which calibration is performed over multiple datasets acquired with different image contrast and phase.
Hall D Tuesday 13:30-15:30
1334. Control of Effective TE for 3D Fast Spin Echo – Image Quality Implications
Reed F. Busse1
1GE Healthcare, Madison, Wisconsin, USA
Fast Spin Echo sequences acquire multiple echoes in a train, thus k-space is filled with echoes of varying TE. This produces signal modulation in ky-kz space, the pattern of which is determined by view order. For 3D-FSE, view ordering is highly flexible. This work introduces a number of novel view ordering strategies and explores the image quality implications of each. It is found that ordering views to produce monotonically varying TEs, with “effective TE” controlled by adjusting the echo train length produces images with the fewest artifacts and highest apparent resolution and contrast.
1335. Pulse Sequence Programming with Shared Components: An Open Source Approach
Jeremy F. Magland1, Walter R.T. Witschey II2
1University of Pennsylvania Medical Center, Philadelphia, Pennsylvania, USA; 2University of Pennsylvania, Philadelphia, Pennsylvania, USA
Improvements in pulse sequence methodology as well as advances in scanner hardware continue to open new possibilities for magnetic resonance imaging. However, this growth is accompanied by increased complexity in the task of pulse sequence programming. As a result there is a growing need to share complex pulse sequences between laboratories and across institutions. Scanner incompatibilities and software complexities can hamper this kind of dissemination. Here we present a pulse sequence development paradigm that seeks to solve these problems by simplifying the pulse program implementation procedure and the act of sharing sequences between laboratories and institutions via scanner-independent, open source sequence development software.
1336. Efficient Pulse-Sequence Simulation Using Variable Subsampling and Decomposition
William R. Overall1, John M. Pauly1
1Stanford University, Stanford, California , USA
A new technique for complete pulse-sequence simulation is described which takes advantage of temporal sparsity and redundancy in MR pulse sequences to greatly reduce computation time without additional a priori assumptions. Using this technique, a wide range of synthetic images can be computed in minutes. Additionally, strategies for decomposing sequence-simulation tasks in order to improve computational efficiency are described. The algorithm is implemented in a flexible, graphical simulation environment that is available online.
1337. The Optimal Acquisition Strategy for Exponential Decay Constants Estimation
Roman Fleysher1, Lazar Fleysher1, Oded Gonen1
1NYU School of Medicine, New York, USA
In this report, chi^2 fitting of multi-point data is used to demonstrate that the most efficient acquisition strategy for estimating relaxation constant of (mono-) exponentially decaying signals is a two-point scheme. We also conjecture that the smallest coefficient of variation of the decay constant achievable in such experiment is 3.6 times larger than that in the image intensity obtained by averaging N acquisitions with minimal exponential weighting.
1338.
Quantifying the Signal to Noise Ratio Benefit of Apodization
by Sampling Density Design:
A
Demonstration with Sodium MRI of the Human
Brain
Robert Wayne Stobbe1, Christian Beaulieu1
1University of Alberta, Edmonton, Canada
For low signal applications like sodium MRI, Gibbs' ringing can degrade image quality. Apodization by sampling density design offers a significant signal to noise ratio benefit over apodization by post-acquisition filtering. This advantage is quanitifed and demonstrated with sodium imaging of the human brain.
1339. Variable Density 3D Shells Acquisition with an Increased FOV
Yunhong Shu1, Matthew Bernstein1
1Mayo Clinic, Rochester, Minnesota, USA
Sometimes 3D non-Cartesian acquisitions like shells require an increased imaging FOV. To keep the spatial resolution constant, the acquisition time needs to be increased roughly as the cube of FOV, which is not acceptable for time-constrained applications. The proposed solution uses a variable density scheme which samples only the central shells more densely, while undersampling the outer shells. Here we demonstrate that variable density sampling for a 3D shells acquisition can increase the imaging FOV with less acquisition time penalty compared to a fully-sampled acquisition strategy. Previously it was shown that the undersampling artifacts are tolerable, especially for contrast-enhanced applications.
1340. Generalized Density Weighted Imaging
Marcel Gutberlet1, Oliver Marcel Geier2, Dietbert Hahn1, Herbert Köstler1
1Universität Würzburg, Würzburg, Germany; 2Rikshospitalet University Hospital, Oslo, Oslo, Norway
A new method "generalized density weighted imaging" is presented that allows simultaneously to optimize the SNR and the SRF for sequences with non – constant magnetization (SR sequences, EPI, FSE) during data acquisition.
1341. Design Metrics for Data Undersampling and Weighting Strategies
James G. Pipe1
1Barrow Neurological Institute, Phoenix, Arizona , USA
There is great interest in undersampled data collection, particularly for Non-Cartesian MRI, as a means to decrease scan time at the minimal expense of aliasing the high spatial frequencies. This work presents a framework in which both relative SNR and aliasing may be compared between different trajectories and data weighting schemes (necessary for nonuniformly sampled data). Ultimately this will aid in both designing and choosing specific trajectories and data-weighting methods.
1342. An Improved Analytical Solution for Variable Density Spiral Design
Tiejun Zhao1, Youngxian Qian2, Yik-Kiong Hue2, Tamer S. Ibrahim2, Fernando Boada2
1Siemens Medical Solutions, USA, Pittsburgh, Pennsylvania, USA; 2University of Pittsburgh, Pittsburgh, Pennsylvania, USA
An improved analytical solution was presented for designing the variable density spiral trajectory. The maximum slew rate of the trajectory generated from the new method was tested for different k-space oversampling factors and varied spiral segments. Compared to the previous method, the new spiral trajectory design method eliminated the slew rate overflow around the k-space center and only slightly increased the total spiral readout duration. The phantom results showed an improved imaging quality for the images acquired with trajectories generated from the new method.
1343. Comparison of Three Radial Trajectories for Highly Time-Resolved, Short TE Imaging
Kuan J. Lee1, Jim M. Wild1
1University of Sheffield, Sheffield, UK
In certain applications, e.g. Hyperpolarized 3He of the lung, both good time resolution and short TE is required to obtain dynamic information in the presence of field inhomogeneities. Symmetric radial, asymmetric echo radial and Quark trajectories are compared to discover (i) which trajectory recovers most signal in the least time under conditions of short T2 and off-resonance (ii) which trajectory gives fewest artifacts when angularly undersampled for increased time resolution. Our simulations show that Quark gives higher time-resolution than symmetric radial, and recovers signal better than asymmetric echo radial. Therefore we conclude Quark represents a good compromise.
1344.
Temporal Stability of 3D-PR Based on Multidimensional Golden
Means: Simulation and Implementation
Rachel Wai-chung Chan1, Donald Bruce Plewes2
1University of Toronto, Toronto, Canada; 2Sunnybrook Health Sciences Centre, Toronto, Canada
Adaptive sampling of k-space allows images in dynamic contrast-enhanced MRI to be reconstructed at various spatial and temporal resolutions from the same dataset. A golden-angle radial k-space sampling scheme achieves this flexibility in-plane with samples incremented by the golden angle. We extend this method to 3D Projection Reconstruction (3D-PR) using multidimensional golden means, which are derived from modified Fibonacci sequences. Using both simulations and experiments, we show that the golden 3D-PR approach has improved temporal stability compared to conventional 3D-PR.
1345. Theoretical and Experimental Aspects of Time Shared Sweep Excitation Using HSn Pulses
Djaudat Idiyatullin1, Curt Corum1, Steen Moeller1, Michael Garwood1
1CMRR, University of Minnesota, Minneapolis, Minnesota, USA
An MRI method called SWIFT (SWeep Imaging with Fourier Transform), has significant benefits for studying objects with ultra fast spin-spin relaxation rates. SWIFT uses swept RF excitation and virtually simultaneous signal acquisition in a time-shared mode. This work investigates frequency-modulated pulses of the HSn family as excitation pulses and how the excitation property of pulses varies with the introduction of the gaps needed for acquisition. The quality of the resulting images highly depends on the fidelity of excitation profile, which can be maximized by oversampling the pulse function. For illustration, 3D SWIFT images of human head are presented and discussed.
1346.
Comparison of Simple and B1-Compensated Spin-Lock
Preparation Techniques at Strong B0 Magnetic
Field Inhomogeneities
Petros Martirosian1, Michael Deimling2, Berthold Kiefer2, Nina Franziska Schwenzer1, Fritz Schick1
1University of Tübingen, Tübingen, Germany; 2Siemens Medical Solutions, Erlangen, Germany
The purpose of this work was a comparison between a simple Spin-Lock (SL) and a B1-compensated SL preparation technique at strong B0 magnetic field inhomogeneities. The behavior of the spin magnetization is described by the rotating matrix formalism. Measurements with a water phantom were performed at 1.5 T using a SL magnetization-prepared gradient echo sequence. Experimental and theoretical results show that B1-compensated SL pulse preparation works well if the SL pulse amplitude is much higher than static local field inhomogeneities. In body regions with strong B0-field inhomogeneities, the simple SL pulse is advantageous for acquisition of artifacts-free images.
1347. Double Inversion Recovery 3D FSE with 2D-Centric Encoding
Reed F. Busse1, Daniel W. Rettmann2, H Glenn Reynolds3, Sterling C. Johnson4, Howard A. Rowley4
1GE Healthcare, Madison, Wisconsin, USA; 2GE Healthcare, Rochester, Minnesota, USA; 3GE Healthcare, Boston, Massachusetts, USA; 4University of Wisconsin, Madison, Wisconsin, USA
Double Inversion Recovery (DIR) nulls CSF and white matter to better visualize gray matter and white matter lesions. A 3D FSE readout with centric view ordering minimizes TE and maximizes SNR, but may result in blurring. A new 2D-centric view ordering for 3D FSE is compared to conventional 1D-centric. Variable flip refocusing is also examined, exploring the trade offs between resolution, SNR and scan time enabled by varying the minimum refocusing flip angle α min) and echo train length. It is found that 2D-centric view ordering results in higher resolution than 1D-centric and resolution can be further enhanced by reducing α min.
1348. Adiabatic Selective Excitation in Single-Slab 3D Turbo Spin Echo Imaging
Jaeseok Park1, John P. Mugler III2
1Siemens Medical Solutions, Erlangen, Germany; 2University of Virginia, Charlottesville, USA
Spatially selective single-slab three-dimensional (3D) turbo spin echo (SE) sequence has been recently developed to increase imaging efficiency employing a highly selective excitation radio-frequency (RF) pulse, very short non-selective refocusing pulses, and variable low flip angles with long echo trains. Despite the enhanced imaging efficiency, this sequence is sensitive to spatially varying B1 amplitude, in particular, at high field, generating non-uniform signal-intensity or contrast over the field-of-view. The purpose of this work is to develop a version of single-slab 3D turbo SE sequence less prone to B1 inhomogeneity without compromising the imaging efficiency using composite adiabatic selective excitation.
1349. Optimized MRI Gradient Waveforms for Acoustic Noise Reduction
Marcel Segbers1, Carlos V. Rizzo S. 1, Hendrik Duifhuis1, Hans Hoogduin1
1University Medical Center Groningen, University of Groningen, Groningen, Netherlands
The noise produced by MR scanners is a major concern. A framework is presented that allows the construction of trapezoidal gradient waveforms that inherently damp two resonance frequencies of the gradient coil. An example is provided which shows a 9 dB reduction in sound pressure level.
Flow Quantification Methods & Applications
Hall D Tuesday 13:30-15:30
Phase-Contrast MRI
Tino Ebbers1, Henrik Haraldsson1, Petter Dyverfeldt1, Andreas Sigfridsson1, Marcel Warntjes1, Lars Wigström1
1Center for Medical Imaging Science and Visualization (CMIV) and Linköping university, Linköping, Sweden
Phase-contrast magnetic resonance imaging has the ability to accurately measure velocities. Unwanted spatially varying phase offsets caused by eddy currents are often corrected by estimating a linear fit of the phase values from either a number of manually defined areas or by semi-automatic detection of stationary tissue. In many applications a linear model is no longer sufficient to describe the phase offset variations. In this study we investigate the variation of the phase offset over the complete 3D volume of interest and propose a completely automated correction approach based on a weighed fit of a high order polynomial.
1368. Velocity Mapping in Highly Stenotic Tubes by the Use of Short-Echo Spiral Acquisitions
Anders Nilsson1, Einar Heiberg2, Freddy Ståhlberg1, Karin Markenroth Bloch3
1Lund University, Lund, Sweden; 2Lund University Hospital, Lund, Sweden; 3Philips Medical Systems, Lund, Sweden
Measurement of peak-velocities in stenotic blood vessels is important in order to estimate the severity of occlusions. As stenotic vessels often present complex flow patterns, lower echo times is desirable. We have evaluated a spiral phase-contrast gradient echo (PC-GRE) sequence together with a conventional cartesian PC-GRE sequence and a k-t BLAST accelerated cartesian GRE sequence (k-t factor of 5) in a stenotic phantom, with respect to flow volume and peak-velocity. The results show that the spiral sequence can be used to depict both flow volume and peak-velocity correctly in shorter acquistion times than a k-t BLAST accelerated gradient sequence.
1369. Efficient Data Acquisition for MR Doppler
Daeho Lee1, Adam Bruce Kerr1, Juan Manuel Santos1, Bob Sueh-Chien Hu2, John Mark Pauly1
1Stanford University, Stanford, USA; 2Palo Alto Medical Foundation, Palo Alto, USA
MR Doppler is a technique that provides real-time imaging of the velocity profile of blood flow analogous to Doppler ultrasound imaging. It provides a mechanism for quickly interrogating valvular flow characteristics, either for identifying valvular stenoses or regurgitant flow. To detect peak velocity of patients, velocity field of view is often required to be in the range -4m/s ~ +4m/s, which results in a lower spatial resolution. We present a flexible design method that makes it easier to realize the desired k-space trajectory in a time-optimal way and this can be used to achieve higher spatial resolution.
1370. Velocity Contour Mapping for Rapid Practical Flow Examination
Ludovic de Rochefort1, Thanh Nguyen1, Pascal Spincemaille1, Yi Wang1
1Weill Medical College of Cornell University, New York, New York, USA
Flow-selective pulses have been proposed using velocity encoding gradients between several RF excitations. Here, we describe a technique to obtain a single image containing velocity isocontours encoded in signal amplitude by tagging in the velocity domain. This method, that provides an original way to look at velocity, is validated on phantom and applied in vivo.
1371. Optimized Pre-Processing Strategy for the Correction of Gradient Field Inhomogeneities in 3D-PC-MRI
Daniel Giese1, Jelena Bock1, Aurelien Stalder1, Ramona Lorenz1, Michael Markl1
1University Hospital Freiburg, Freiburg, Germany
Phase Contrast MRI is widely used to assess and quantify blood flow and tissue motion. In particular for applications analyzing 3D blood flow within large volumes gradient non-linearities can introduce substantial errors in the encoded velocities. This study presents an optimized pre-processing strategy to retrospectively correct velocity data. Application in phantom studies demonstrated its correction impact on velocity encoded data and confirmed the necessity for a systematic correction of the effect of gradient imperfections on 3D Phase Contrast measurements.
1372.
Effect of Voxel Size, Image Orientation, VENC and TE on Phase
Contrast Measurements in
Turbulent Stenotic Jets
Kieran R. O'Brien1, Brett R. Cowan1, Andreas Greiser2, Alistair A. Young1
1University of Auckland, Auckland, New Zealand; 2Siemens, Erlangen, Germany
Aortic Stenosis may lead to turbulent jets of more than 400cm/s. Often these jets suffer from significant errors in peak velocity and flow due to intravoxel dephasing. The factors affecting the accurate measurement of these high velocity turbulent jets remains unclear. We have investigated, using up-to-date gradient hardware, the influence of TE, voxel size, velocity encoding and image orientation relative to the jet . All of which have previously been found to cause signal loss and intravoxel dephasing in high velocity turbulent jets.
1373. Clinical Evaluation of Aortic Coarctation with 4D Flow MR Imaging
Michael D. Hope1, Thomas A. Hope1, Karen Ordovas1, Alison Meadows1, David Saloner1, Gautham P. Reddy1, Marc T. Alley, Charles B. Higgins1
1UCSF, San Francisco, California , USA
In order for 4D Flow to be a clinically viable tool for evaluation of aortic coarction, it must be reasonably fast and generate reliable blood flow data, as well as offer advantages over the standard MR protocol. We address these first two objectives using parallel imaging in a direct comparison of 4D Flow and 2D phase contrast data in patients. With regard to its advantages, 4D Flow allows does not require prospective placement of 2D planes for PC acquisition, and offers unique visualization of flow data that is not available by traditional 2D PC imaging.
1374.
Late Consequences After Surgery of Left Sided Diaphragmatic
Hernia in Children: Lung Volumes
and Blood Flow
Nasreddin Abolmaali1, Knut Götzelt2, Arne Koch1, Christian Vogelberg3, Michael Laniado, Gabriele Hahn2
1OncoRay, TU Dresden, Dresden, Germany; 2Pediatric Radiology, Dresden, Germany; 3Pediatric Pulmonology, Dresden, Germany
To compare pulmonary arterial and cardiac findings in children operated for left-sided diaphragmatic hernia in comparison with healthy children. At least six years after surgery, patients received a clinical interview, pulmonary function testing and echocardiography. Furthermore, all children received MRI examinations for functional cardiologic evaluation and velocity encoded flow measurements of pulmonary arteries. Examinations other than MRI revealed neither pathologic findings nor significant differences between volunteers and patients. On the other hand, MRI showed significantly reduced cardiac volumes, while the heart rate was increased. The flow parameters of the left pulmonary artery in patients were reduced and pulmonary blood flow was shifted from the left to the right main pulmonary artery. These findings indicate persistent pulmonary hypotrophia on the operated side.
1375. Accurate Quantification of Aortic Regurgitation with PC-MRI
Meera Sekar1, John Sheehan, Jennifer Berliner, Issam Mikati, James Carr2
1Northwestern University, Chicago, USA; 2Northwestern University, USA
This retrospective study is a comparison of MRI quantification of aortic regurgitation at varying anatomic levels in relation to the aortic valve with Echocardiographic quantification.
1376. Hemodynamic Response to Exercise in Small Abdominal Aortic Aneurysms
Adam Sebastian Tenforde1, Christopher P. Cheng1, Kelly Y. Suh1, Andrea S. Les1, Ronald L. Dalman2, Robert J. Herfkens1, Charles A. Taylor1
1Stanford University, Stanford, USA; 2Stanford University Medical Center, Stanford, USA
Abdominal Aortic Aneurysm (AAA) is a vascular disease whose course may be altered by exercise. Six male patients with small AAA were compared to eight healthy control patients while performing lower limb exercise on an MR-compatible bicycle. PC-MRI data was collected at supraceliac, infrarenal, and mid-aneurysm positions on a GE 0.5T MRT. Measured outcomes include average mean flow and oscillatory flow index (OFI). AAA patients had similar values of average mean flow to control patients during rest and exercise, and AAA patients reduced OFI during exercise. These results suggest mild exercise may reduce adverse hemodynamic conditions in AAA disease.
1377. Vortex Ring Formation in Diastolic Dysfunction: Phase Contrast MRI of Left Ventricular Filling
William Sean Kerwin1, David Owens1, Jean Hertzberg2, Robin Shandas2, Edward Gill1
1University of Washington, Seattle, Washington, USA; 2University of Colorado, Boulder, Colorado , USA
We developed a method for viewing complex 3D vortex ring formation in the left ventricle during filling using CINE phase-contrast MRI. We evaluated the potential of the method by comparing vortex ring formation in 5 subjects with diastolic dysfunction to 5 normal subjects. Two distinct vortices were observed during filling, with the first appearing significantly weaker and delayed in subjects with diastolic dysfunction. We conclude that vortex ring characteristics observed with this method may be valuable for diagnosis and staging of diastolic dysfunction.
1378.
Breathhold Time-Resolved Three-Directional MR Velocity
Mapping of Aortic Flow in Patient Follow-Up
After Aortic Valve-Sparing
Surgery
Xin Liu1, Peter Weale, Randall Ramsay1, Aya Kino1, Gert Reiter, Karin Dill1, James Carr1
1Northwestern University, Chicago, Illinois, USA
To evaluate the potential of breathhold time-resolved three-directional MR velocity mapping for the follow-up in patients after aortic valve-sparing surgery, 13 patients with valve-sparing surgery as well as 10 healthy volunteers and 12 patients with ascending aortic aneurysm were evaluated by time-resolved three-directional MR velocity mapping. Aortic laminar flow, turbulent flow, and the presence of vertical flow in the sinuses of Valsalva were analyzed. Breathhold time-resolved three-directional MR velocity mapping allows for the detection of flow patterns in aortic root and ascending aorta. Normal laminar flow in the ascending aorta and vertical flow in the sinuses of Valsalva can be restored in patients after aortic valve-sparing surgery. Evaluation of aortic flow patterns using MR velocity mapping may be a useful indicator for outcome of patients with aortic valve-sparing surgery.
1379. Non-Contrast-Enhanced MRA of the Renal Vasculature with the BSSFP Dixon Method
Randall B. Stafford1, 2, Mohammad Sabati1, 2, Michael J. Haakstad2, M Louis Lauzon1, 2, Houman Mahallati1, 2, Richard Frayne1, 2
1University of Calgary, Calgary, Canada; 2Foothills Medical Centre, Calgary Health Region, Calgary, Canada
Renal artery stenosis is a major cause of renal insufficiency. Gadolinium-based MR contrast agents have been linked to nephrogenic systemic fibrosis in patients with renal insufficiency. The purpose of this research was to collect non-contrast-enhanced MR angiograms of the renal arteries using the balanced steady-state free precession Dixon method in healthy volunteers. The results showed that the technique produced images with good fat-suppression and strong arterial vessel conspicuity. We conclude that this technique is suitable for non-contrast-enhanced renal MR angiograms.
Hall D Tuesday 13:30-15:30
Alex Xuexin Li1, 2, Mojmir Suchy, 12, Craig K. Jones1, Ravi S. Menon1, 2, Robert H.E. Hudson2, Robert Bartha1, 2
1Robarts Research Institute, London, Canada; 2The University of Western Ontario, London, Canada
The linear relationship between temperature and bound water chemical shift of MRI-PARACEST contrast agent might be used to measure temperature in-vivo. Temperature maps with standard deviations <1 C were acquired at 9.4 Tesla in phantoms containing A) an aqueous solution of 10 mM Eu3+-DOTAM-Gly-Phe, B) 5% bovine serum albumin with 15 mM Eu3+-DOTAM-Gly-Phe, and C) mouse brain tissue with 4 mM Eu3+-DOTAM-Gly-Phe.
1438. Modulated CEST, Using Cross-Correlation to Measure the CEST Effect
Thijs Hendrix1, 2, Klaas Nicolay1, Rolf Lamerichs2
1Eindhoven University of Technology, Eindhoven, Netherlands; 2Philips Research, Eindhoven, Netherlands
A new method is presented to asses the CEST effect. This modulated CEST uses the correlation between the signal change and the RF saturation paradigm. The advantages are that it can be executed real-time. The correlation method will detect the CEST contrast more reliably than averaging when additional signal changes occur, caused by motion or flow artefacts.
1439. CEST in the Presence of MT
Kimberly L. Desmond1, Greg Jan Stanisz1, 2
1University of Toronto, Toronto, Canada; 2Sunnybrook Health Sciences Centre, Toronto, Canada
This work explores the validity of the assumption that chemical exchange saturation transfer (CEST) and magnetization transfer (MT) effects are decoupled, with relevance to in-vivo applications involving amide proton transfer (APT) and PARACEST contrast agents. We compared the results of a two-pool compartmental model containing only a CEST component, to a three-pool model which also included MT parameters. It was found that a description of the CEST effect in terms of asymmetry alone may result in its underestimation in the presence of MT, leading to errors in the calculation of physiological quantities including pH and solute concentration.
Alex Xuexin Li1, 2, Craig K. Jones1, Mojmir Suchy, 12, Ravi S. Menon1, 2, Robert H.E. Hudson2, Robert Bartha1, 2
1Robarts Research Institute, London, Canada; 2The University of Western Ontario, London, Canada
A four-pool model based on the modified Bloch equations with exchange terms for the proton exchange processes of a PARACEST agent in biological systems was used to predict the optimal saturation pulse duration for maximum on-resonance PARACEST contrast in-vivo. This predication was verified experimentally by imaging aqueous and bovine serum albumin (BSA) phantoms containing Tm3+-DOTAM-Gly-Lys.
1441.
Evaluation of Hyperoxic Gas Induced δR1 and δR2*
as MRI Biomarkers of Tissue Oxygenation Status in
Human Subjects
James P B O'Connor1, 2, Josephine H. Naish1, David L. Buckley1, Alan Jackson1, John C. Waterton1, 3, Yvonne Watson1, Giovanni A. Buonaccorsi1, Deirdre M. McGrath1, Sue Cheung1, Samantha J. Mills1, Gordon C. Jayson2, Geoffrey J M Parker1
1University of Manchester, Manchester, UK; 2Christie Hospital NHS Trust, Manchester, UK; 3AstraZeneca, Macclesfield, UK
Differences in tissue longitudinal relaxation rate (R1) and effective transverse relaxation rate (R2*), induced 100 % oxygen and carbogen inhalation were evaluated in ten healthy subjects. Significant reductions in R1 were demonstrated in spleen, liver and renal cortex, following oxygen and carbogen inhalation. Significant increase in R2* was observed in all organs following carbogen inhalation, an opposite effect to that observed in many studies of tumour pathophysiology. R2* was not changed by breathing 100 % oxygen. The R1 and R2* changes represent distinct but complementary mechanisms of tissue contrast and this study offers further insight into their physiological mechanisms.
1442.
Water Diffusion and Magnetization Transfer in Normal and
Pathologic Lumbar Disc
Meritxell Garcia1, Thomas Egelhof2, Klaus Scheffler3, Oliver Bieri3
1University Hospital Erlangen, Erlangen, Germany; 2University Hospital Basel, Switzerland; 3University of Basel, Switzerland
Diffusion has been shown to be sensitive to intervertebral disc degeneration and pathologies but quantification is often hampered by global motion and susceptibilities. In this work, we investigate possible correlations between the magnetization transfer ratio (MTR) imaging and apparent diffusion constamts (ADC) to yield a possible alternative characterization method of disc diseases that relate to diffusion. ADC and MTR is compared between normal appearing discs of young healthy volunteers and patients. High correlations between MTR and ADC values were found suggesting that changes in water diffusion for intervertebral discs are also reflected in MT measurements.
1443. Delta Relaxivity Enhanced MR (DreMR): Theory of T1-Slope Weighted Contrast
Jamu K. Alford1, Brian K. Rutt2, Timothy J. Scholl1, William B. Handler1, Blaine A. Chronik1
1The University of Western Ontario, London, Canada; 2Robarts Research Institute, London, Canada
Delta relaxivity enhanced MR (dreMR) is a novel MR method for producing image contrasts related to the magnetic field dependence of tissue relaxation rates. Applications include cellular/molecular MRI, where dreMR may significantly increase the detection sensitivity/specificity to in-vivo target molecules. Molecular specificity is obtained by accessing the dramatic change in relaxivity slope, seen upon molecular binding of activatable contrast agents, as the source of contrast. The dreMR double inversion recovery method as well as the necessary hardware requirements to perform dreMR in a clinical MR scanner will be discussed.
1444. Modulation of Tumour R1: A Novel Biomarker of Oxygenation Status
James P B O'Connor1, 2, Alan Jackson1, Giovanni A. Buonaccorsi1, Yvonne Watson1, Sue Cheung1, Gordon C. Jayson2, Geoffrey J. Parker1
1University of Manchester, Manchester, UK; 2Christie Hospital NHS Trust, Manchester, UK
Inhalation of 100% oxygen has been shown to increase normal tissue R1. We describe modulation of tumour R1 in five patients with advanced solid tumours. Significant increase in R1 was measured on oxygen inhalation(p<0.001; group range 0.0087-0.0526 s-1). This suggests that the technique shows promise for producing novel biomarkers of tumour oxygenation status.
William R. Overall1, John M. Pauly1
1Stanford University, Stanford, California , USA
We propose a new technique for positive-contrast visualization of small volumes of paramagnetic agents using fluctuating-equilibrium steady state in conjunction with opposing refocusing gradients. The combination of these two techniques provides better background suppression than either one alone, and the use of a dual-equilibrium steady state allows for concurrent acquisition of an anatomic reference. Detection limits of the technique are explored theoretically and experimentally, and images demonstrate improved detection when compared to previous techniques.
1446. Positive Contrast with Therapeutic Iron Nanoparticles for in Vivo Follow Up at 4.7T
Monica Sigovan1, Misara Hamoudeh2, Achraf Al Faraj1, Hatem Fessi2, Emmanuelle Canet-Soulas1
1Université de Lyon, Université Lyon1, Creatis-LRMN, UMR CNRS 5220, U630 INSERM, INSA-Lyon, Lyon, France; 2Université de Lyon, Université Lyon1, LAGEP, UMR CNRS 5007, Lyon, France
Magnetic nanoparticles (MNP) are expected to enable the inclusion of specific drugs in order to combine a therapeutic action to the MRI diagnostic approach. The importance of positive contrast techniques is the improved sensitivity for iron detection compared to standard T2* methods. We report the application of the “white marker” technique at a relatively high field of 4.7 T in a preliminary in vitro and in vivo study using a novel MNP system designed for local therapy. Compared to the standard gradient echo the method shows the ability to discern between high concentrations. The combined negative and positive contrast protocol allowed a good characterisation for the in vivo localization of MNP and their follow up.
1447. Positive Contrast Fe Nano-Particle Imaging of Mouse Brain Vasculature with SWIFT
Curtis Andrew Corum1, Geoffrey L. Curran2, Deepali Sachdev1, Djaudat Idiyatullin1, Steen Moeller1, Michael Garwood1
1University of Minnesota, Minneapolis, Minnesota, USA; 2Mayo Clinic College of Medicine, Rochester, Minnesota, USA
We demonstrate initial results with positive contrast enhancement in an in-vivo wild type mouse brain, after bolus I.V. injection(s) of the mono-crystalline ion oxide nano-particle solution1 MION-47. MION-47 has both R1 and R2* relaxivity, but the R2* relaxivity typically dominates at high concentrations and/or high fields.Utilizing SWIFT (SWeep Imaging with Fourier Transform) we show high field, high concentration R1 induced positive T1 contrast at 5mg/kg and 20 mg/kg MION-47 doses.
1448. Visualization of Inhomogeneous Local Magnetic Field Gradient Due to Susceptibility Contrast
HyungJoon Cho1, Seungoh Ryu2, Jerry L. Ackerman3, Yi-Qiao Song2
1Memorial Sloan Cancer Center, New York, New York, USA; 2Schlumberger Doll Research, Cambridge, Massachusetts, USA; 3Martinos Center, Department of Radiology, Massachusetts General Hospital, Cambridge, Massachusetts, USA
We visualize inhomogeneous local magnetic field (internal magneticfield) gradient arising from susceptibility contrast between anarray of cylindrical glass tubes (solid matrix) and surroundingwater (pore fluids) in a uniform applied magnetic field. MRIwas performed to determine the spatially resolved decay rates due to diffusion in the internal magneticfield. These rates are shown to be directly proportional to thelocal gradient strength obtained from theoretical calculations. Wealso spatially resolve the interference pattern ofthe cross-terms between internal and external pulsed fieldgradient (PFG) along different PFG orientations and extractcorresponding cross-terms in this model system. This workdemonstrates a simple yet representative case for visualizing thelocal susceptibility induced magnetic fields in porous media.
1449.
Combined Positive Contrast and Relaxation in the Rotating
Frame for Molecular Imaging of In-Vivo
SPIO Labeled Cells
Ovidiu Cristian Andronesi1, 2, Dionyssios Mintzopoulos1, 2, Meenu Kesarwani1, Laurence Rahme1, Aria A. Tzika1, 2
1Massachusetts General Hospital and Shriners Burn Hospital, Harvard Medical School, Boston, Massachusetts, USA; 2Athinoula A. Martinos Center for Biomedical Imaging, Boston, Massachusetts, USA
We present a refined MRI method for positive contrast (PC) molecular imaging of superparamagnetic iron-oxide (SPIO) labeled macrophages to visualize infection in-vivo and monitor its progression. Existing negative and positive contrast methods for SPIO detection have difficulties due to quantification and specificity limitations, respectively. We combine existing positive contrast schemes with relaxation in the rotating frame to improve the specificity for SPIO. Our results on phantom, and test and control mice suggest that this combination can be used to fine-tune the level of contrast and to quantify the SPIO and macrophage accumulation at the infection site.
Sherif R. Fahmy1, Nicole Mascheri1, Tanja Paunesku1, Gayle Woloschak1, Debiao Li1, Todd Parrish1
1Northwestern University, Chicago, Illinois, USA
An off-resonance preparation pulse is used to selectively dephase spins in the vicinity of a super-paramagnetic tracer. Short echo times reduce T2* effects outside of the frequency range being saturated. The signal from background tissue remains intact as compared to positive contrast techniques for tracer detection.
1451. Off-Resonance Projection Imaging of USPIO-Enhanced Bone Marrow
Carsten Warmuth1, Michael Reinhardt2, Hanns-Joachim Weinmann2, Robert Krieg1
1Siemens Medical Solutions, Erlangen, Germany; 2Bayer Schering Pharma, Berlin, Germany
Ultrasmall superparamagnetic iron-oxide (USPIO) particles are taken up by healthy bone marrow. With an off-resonance spin echo sequence, projection images of the bone marrow comparable to scintigraphy or X-Ray acquisitions can be obtained.
1452. Assessing the Detection Sensitivity of Iron Loaded Cells in Spoiled Gradient Echo Imaging
Xavier Helluy1, Wing Chow2, daniel Haddad1, R Ebert1, M Weber1, P M. Jacob1
1University of Würzburg, Würzburg, Germany; 2University of Toronto, Toronto, Canada
Susceptibility inhomogeneities induced by iron particles affect the magnitude and phase of gradient echo images, facilitating the detection of iron-loaded cells with MR. The goal of this study is to quantify the iron detection sensitivities of spoiled gradient echo magnitude and phase images and evaluate the possible gain in iron detection sensitivity when combining phase and magnitude information. To this end extended statistical analysis of numerical simulations and experimental measurements on iron-loaded cells dispersed in gels have been conducted, and various image filters evaluated.
Nicole Mascheri1, Zhuoli Zhang1, Tanja Paunesku1, Gayle Woloschak1, Rohan Dharmakumar1, Debiao Li1
1Northwestern University, Chicago, Illinois, USA
Inflammatory activity in atherosclerotic lesions is an indicator of plaque vulnerability. MRI is capable of detecting inflammation through accumulation of ultrasmall superparamagnetic iron oxide nanoparticles (USPIO) in macrophage cells in the plaque. The purpose of this work was to investigate the application of a recently developed positive contrast imaging method, termed Fast Low-Angle Positive contrast Steady-state free precession (FLAPS) imaging, to USPIO-labeled macrophages at clinical field strength and resolution. The dependence of positive contrast on flip angle and concentration of labeled cells was also evaluated.
Elvira Mehlin1, Stefan Kirsch1, Peter Bachert1
1German Cancer Research Center (DKFZ), 69120 Heidelberg, Germany
The CRAZED sequence permits detection of signal generated by intermolecular double–quantum coherence (iDQC). When applied to 1H MRI, a novel type of contrast is obtained, in particular, a positive contrast when superparamagnetic iron oxide particles (SPIOs) are present. We demonstrate that the high T1–relaxivity of SPIOs can cause spurious signal in CRAZED MRI which cannot be attributed to iDQC.
Peter Roland Seevinck1, Jan-Henry Seppenwoolde1, Jaco Zwanenburg1, Chris J.G. Bakker1
1University Medical Center, Utrecht, Netherlands
Quantitative assessment of the radionuclide biodistribution is of major importance for dosimetry in transcatheter hepatic arterial embolization with radioactive holmium-loaded microspheres (HoMS). In this work, we demonstrate that T2*-based quantification of holmium-loaded microspheres can be done best using multiple gradient echo sampling of FID (MGEFID) rather than SE (MGESE) signals. MGESE caused a severe underestimation of the integral HoMS dose present, specifically at higher concentrations, which can be attributed to the diffusion sensitivity of MGESE, as was shown by in vitro. MGEFID strongly reduced the underestimation of HoMS in ex vivo rabbit livers.
1456. Modeling CPMG Behavior in a Realistic Magnetically-Heterogeneous Tissue Matrix
Nilesh R. Ghugre1, 2, John C. Wood1
1Childrens Hospital Los Angeles, Los Angeles, California , USA
R2* and single echo R2 have been successfully calibrated for hepatic iron concentration. Despite several advantages of CPMG sequences, their implementation in quantifying clinical iron burden has been difficult. They are complicated by anomalous non-exponential signal decay characteristics and R2-iron calibrations may also vary with echo spacing t). We employed a realistic tissue geometry to interrogate the inner mechanisms of CPMG behavior in the presence of magnetic inhomogeneities, via computational modeling. Model-predicted R2t behavior was consistent with previous findings. Moreover, predicted R2-iron relationships at specified t were in confirmation with published in vivo calibration curves. A validated computational model will complement existing theoretical predictions in understanding the underlying biophysics of proton-iron interaction in an iron-rich tissue matrix.
1457. MR Contrast Media at 7Tesla - Preliminary Study on Relaxivities
Iris M. Noebauer-Huhmann1, Oliver Kraff2, Vladimir Juras1, Pavol Szomolanyi1, 3, Stefan Maderwald2, Vladimir Mlynarik4, Jens M. Theysohn2, Susanne C. Ladd2, Mark E. Ladd2, Siegfried Trattnig1
1MR Centre - Highfield MR, Vienna, Austria; 2Erwin L.Hahn Institute for MRI, Essen, Germany; 3Institute of Measurement Science, Slovak Academy of Sciences, Bratislava, Slovakia; 4EPFL / LIFMET, Lausanne, Switzerland
Contrast media are a powerful tool to improve the diagnostic value of MRI. The relaxivities have been determined for the routinely used magnetic field strengths 1.5T and 3T. Recently, 7 Tesla magnets have been introduced for whole body imaging. In preliminary calculations, the relaxivities r1 and r2 in physiologic saline soluton at 37 C were lower at 7 Tesla than has been described in the literature for lower field strengths. The values decrease with higher contrast agents concentrations. The values for blood were also calculated. The different r1 and r2 relaxivities at 7 Tesla have to be taken into account.
Chung Ki K. Wong1, Jianhui Zhong1
1University of Rochester, Rochester, New York, USA
The decrease of dipolar signal with refocusing pulses of finite duration in a CRAZED-multiecho acquisition was investigated previously. It was found that the rephasing of the dipolar signal during the refocusing pulses in the multiecho sequence depends substantially on the phase of the pulses. In this abstract, the total signal acquired from the multiecho sequence was optimized with the parameters of the sequence. The results show that the attenuation of the signal due to the finite duration of the refocusing pulses can be compensated with a longer t2 and a proper choice of the phase of the pulses.
1459. Cross-Relaxation in Methacrylic Acid-Based Dosimetry Gels
Heather Marie Whitney1, Jim Joers1, John C. Gore1
1Vanderbilt University, Nashville, Tennessee, USA
Radiation dose distributions can be quantified using MRI of irradiated polymer gels The dose response mechanism responsible is believed to involve magnetization exchange between polymer protons and the bulk water in the gel. This study seeks to quantify the contributions of chemical exchange or dipolar cross-relaxation to the dose response mechanism in methacrylic acid-type polymer gels through studies of relaxation properties of the monomers present in the gel system. The results show that dipolar cross-relaxation between water and methacrylic acid monomer may contribute to relaxation through magnetization transfer between the methyl proton and water and one methylene proton and water.
1460.
Synthesis and Physicochemical Characterization
of a New Gd Complex and Its Eu Analogue, Suitable
Bimodal Contrast
Agents for MRI and Optical Imaging
Sophie Laurent1, Luce Vander Elst1, Mélanie Wautier1, Chantal Galaup2, Claude Picard2, Robert Muller1
1University of Mons-Hainaut, Mons, Belgium; 2Université Paul Sabatier, Toulouse, France
The lanthanide complexes of the PMN-tetraacetic acid present thus interesting properties both for MRI and for optical imaging. In fact the Ln-complexes have a relatively good stability in physiological environment. The Gd-complex has a high proton relaxivity and a value of tM which approaches the optimal value required to obtain high relaxivity once the chelate is bioconjugated to macromolecules. Moreover, an efficient energy transfer from the pyridine to the metal occurs in the Eu-complex and the luminescence lifetime of the complex is long enough to avoid the overlapping with biological background.
1461. Compliance Weighted Imaging in MR Elastography
ingolf sack1, Eberhard Siebert1, Uwe Hamhaber1, Dieter Klatt1, Sebastian Papazoglou1, Jürgen Braun1
1Charité - University Medicine Berlin, Berlin, Germany
Compliance weighted imaging (CWI) is introduced that uses ultra-low frequency mechanical excitation. As a result, viscous damping is negligible which allows the approximation of a constant strain-energy averaged over one oscillation period. Using first-derivative data-processing yields a new contrast that is scaled by the inverse stiffness (compliance) of the material. The spatial resolution in CWI-MRE is comparable to that of standard MR images which considerably improves the detail resolution in MRE. CWI-MRE allows mapping the anatomical structure of the brain based on its mechanical properties. The method can further be used to examine mechanical properties of pathologic tissue.
Hall D Tuesday 13:30-15:30
Daniel Henrik Adler1, 2, Sonny Chan, 23, Eric Scott Penner1, 2, J Ross Mitchell1, 2
1University of Calgary, Calgary, Canada; 2Foothills Med Ctr, Calgary Health Region, Calgary, Canada; 3Stanford University, Palo Alto, USA
We describe an automatic medical image registration framework for which all computationally and memory intensive components are implemented on graphics hardware. We accelerate linear and deformable registration 10 to 20-fold on commodity workstations. Accuracy and speed evaluations are performed on neurological clinical data and against other software packages.
1506. Eliminating Functional Localizers Using a Probabilistic Atlas of V1
Oliver Hinds1, Jonathan Polimeni, Mukund Balasubramanian, Graham Wiggins, Florian M. Meise2, Eric L. Schwartz, Bruce Fischl, Lawrence L. Wald, Christina Triantafyllou1
1MIT, Cambridge, Massachusetts, USA; 2Section of Medical Physics, Department of Diagnostic and Interventional Radiology, Mainz University, Germany
Functional localizers simplify the analysis and interpretation of fMRIdata, but cost substantial scantime that could be devoted toinvestigation of other brain function. Here we demonstrate anautomatic method for predicting the boundary of the human primaryvisual cortex (V1) directly from structural scans using aprobabilistic atlas, thus eliminating the need for a V1 localizer. TheV1 boundary was estimated using both fMRI-based localizers and aprobabilistic atlas in six subjects. We found an average distance of less than 7 mm (~2 functional voxels) between theatlas-predicted boundary and the fMRI boundary across subjects.
1507. Development and Application of a Quantitative Water Content Brain Atlas
Veronika Ermer1, Heiko Neeb1, Nadim Jon Shah1, 2
1Research Centre Juelich, Juelich, Germany; 2Institute of Physics, Dortmund, Germany
A series of spoiled gradient echo images with different T2*-weighting (QUTE) was acquired in order to create high-resolution quantitative water maps of the human brain. The resulting maps of several volunteers were transformed to the same stereotactic space and were averaged. A first comparison of the new water content atlas of the human brain with the brain of an MS patient is shown.
1508. Non-Rigid Registration of Histological and MRI Sections for Prostate Cancer Mapping
Jonathan C. Chappelow1, Anant Madabhushi1, John Tomaszewski2, Michael Feldman2, Mark Rosen2
1Rutgers University, Piscataway, New Jersey, USA; 2University of Pennsylvania, USA
Multimodal image registration methods are also under active development for a variety of visualization and diagnostic applications such as image guided surgery and multimodal image fusion for cancer diagnosis and treatment planning. We present a new non-rigid registration method termed COFEMI-TPS for robust alignment of multimodal images, and demonstrate the method for alignment of 26 prostate MRI-histology slice pairs by identifying spatial extent of cancer on MRI by directly mapping histological cancer ground truth to the resulting co-registered MRI. It was observed that the cancer labels on MRI found using our registration method are qualitatively accurate and comparable to expert-determined labels.
Mariano G. Uberti1, 2, Yutong Liu1, Huanyu Dou1, Howard E. Gendelman1, Michael Douglas Boska1
1University of Nebraska Medical Center, Omaha, Nebraska, USA
A methodology for registration of in-vivo MRI to histology using 3D block face histology is presented. Superparamagnetic iron oxide (SPIO) labeled bone-marrow derived macrophages (BMM) were injected intravenously and tracked across the blood-brain barrier (BBB) using T2*-weighted MRI. Histological sections (HS) were obtained and stained with Prussian blue. MRI volumes were registered to HS using the introduced methodology. SPIO-labeled cells were detected by subtraction of Pre MRI minus 7 day Post MRI and compared to stained histology. We found that spatial distribution and cell areas detected with in vivo MRI show a good correlation with histopatology. We conclude that MRI can be used in vivo to track cells migrating across the blood-brain barrier in a mouse model of HIV-1 encephalitis. However, further study is required to determine sensitivity and specificity of the method.
1510. Novel Contour-Based Registration Algorithm for VBM Pre-Processing
Joao M. S. Pereira1, Peter J. Nestor1, Guy B. Williams1
1University of Cambridge, Cambridge, UK
A novel contour based registration algorithm is proposed which uses as input registered grey and white matter segments output by SPM5. The main goal is to improve surface (ventricles and cortex) normalization, which is known to be sub-optimal in SPM. A two-step method is described for each slice: first, subject and target contours are matched and registered by translations using an approximation to elastic registration; then, demon-based forces are used to drive a viscous-fluid registration around the contours. An increased subject/target similarity was achieved for both controls and Alzheimer’s disease patients.
Hall D Tuesday 13:30-15:30
Uwe Hamhaber1, Dieter Klatt1, Bernd Beierbach1, Sebastian Papazoglou1, Maurice Hollmann2, Jörg Stadler3, Ingolf Sack1, Johannes Bernarding2, Jürgen Braun1
1Charité - Universitätsmedizin Berlin, Berlin, Germany; 2Otto-von-Guericke University, Magdeburg, Germany; 3Leibniz Institute for Neurobiology, Magdeburg, Germany
MR elastography allows the quantification of viscoelastic parameters of the living brain at 1.5 T and 3 T. The stronger static magnetic field at 7 T imposes imaging complications by RF resonances and stronger susceptibility artifacts and distinctly different relaxation times of brain tissue. The objective of the presented study is to investigate whether the determination of viscoelastic properties of brain tissue is feasible at 7T and if the reconstructed viscoelastic parameters are comparable with those determined at 1.5 T.
1547. MR Elastography Reveals Tissue Degeneration in Multiple Sclerosis Patients
Jens Thomas Wuerfel1, 2, Bernd Beierbach1, Dieter Klatt1, Uwe Hamhaber1, Sebastian Papazoglou1, Juergen Braun1, Ingolf Sack1
1Charité - University Medicine Berlin, Berlin, Germany; 2University Schleswig-Holstein, Luebeck, Germany
Physicians have been using palpation as an effective clinical examination tool for centuries. However, the brain, an organ that eluded the examination by the palpating hand, has not thus far been accessible in vivo. Magnetic resonance elastography is a unique, non-invasive approach to evaluate the elasticity of soft tissues in vivo. Here, we present cerebral viscoelasticity data from 40 multiple sclerosis patients with relapsing-remitting disease course, as well as from 17 matched healthy controls. Mean shear stiffness was significantly lower in MS patients, exceeding 0.5kPa in comparison with healthy brains. MRE is feasible for application in a routine clinical setting.
1548.
Radiation Force Imaging and HIFU Therapy Monitoring in
Phantom Gels by Means of High Resolution
MR-Elastography – in Vivo
Application to the Rat Brain
Benoit Larrat1, Ralph Sinkus1, Mathieu Pernot1, Jean-François Aubry1, Mickael Tanter1, Mathias Fink1
1ESPCI-CNRS-Univ Paris 7-INSERM, Paris, France
In this study, an HIFU experiment together with an MR-Elastography acquisition is performed inside a 7T scanner. The exact position of the ultrasound focal point is visualized by means of a dedicated motion sensitized sequence. This allows checking the correct positioning of the targeted tissue relatively to the ultrasound beam with low power. The elasticity maps are calculated before and after HIFU treatment, showing significant stiffening at the exact location of the focal spot. In vivo preliminary experiments in the rat brain have also been conducted.
1549.
In Vivo Myocardial MR Elastography: Observation of
Stiffness-Related Shear-Wave Amplitude Variations
Thomas Elgeti1, Jens Rump1, Dieter Klatt1, Uwe Hamhaber1, Sebastian Papazoglou1, Jürgen Braun1, Bernd Hamm1, Ingolf Sack1
1Charité Universitätsmedizin Berlin, Berlin, Germany
Wave amplitude variations of external introduced low frequency shear waves are detected with phase-sensitive MR elastography sequences in human myocardium of 6 healthy volunteers.
1550. MRE of the Eye: Inversion Using a Thin Spherical-Shell Model
Daniel V. Litwiller1, Sung Lee1, Arunark Kolipaka1, Kevin J. Glaser1, Jose S. Pulido1, Richard L. Ehman1
1Mayo Clinic, Rochester, Minnesota, USA
Assessments of ocular, intraocular and orbital rigidity, are currently limited to qualitative assessment by direct palpation, more invasive methods or other conventional methods such as tonometry, which may yield indirect or inaccurate results. Recently, MR Elastography (MRE) has emerged as a promising technique for investigating motion in the eye and other fluid-filled membranes. Because the eye is not a homogeneous solid, however, reconstructing its shear stiffness with existing wave inversion algorithms is not feasible. In this work, we demonstrate that it is possible to interpret MRE images of flexural waves in the eye using direct inversion in a thin spherical-shell model.
Hall D Tuesday 13:30-15:30
Karl Young1, Patrice Weber1, Varanavasi Govindaraju, Khema Sharma, Ammar Darkazanl, Colin Studholme1, Lawrence Hall, Andrew A. Maudsley, Norbert Schuff1
1University of California and VA Medical Center, San Francisco, California , USA
Voxelwise multivariate statistical mapping is introduced for analysis of spectroscopic imaging (SI) data. Applications to experimental SI data demonstrate that multivariate statistical mapping yields greater power for detection of regional metabolite alterations than univariate tests implying that multivariate statistical mapping should be used for SI whenever simultaneous changes of metabolites are expected.
1619.
Semi-Parametric Estimation in Magnetic Resonance
Spectroscopy: Automation of the Disentanglement
Procedure
Hérald Rabeson1, Hélène Ratiney2, Dirk van Ormondt3, Danielle Graveron-Demilly2
1Université Lyon 1, Villeurbanne, France; 2Université Lyon 1, Villeurbanne, France; 3Delft University of Technology, Delft, Netherlands
Magnetic Resonance Spectroscopy (MRS) is a unique tool for non-invasive in vivo detection and quantitation of metabolites. A point of concern is disentanglement of perturbing – macromolecules and lipid - signals from the wanted metabolite signals. Cramér-Rao bounds (CRBs) on the data-points of the perturbing signals are derived. Exploiting the attendant formulae, we show that it is possible to automate – for use in clinics – the disentanglement procedure in the metabolite quantitation algorithm ‘QUEST’, part of the freely available MRS software jMRUI.
1620. MRSTANK - A User-Friendly Yet Powerful Alternative Software Package for Spectral Processing
Marc Rea1
1Charing Cross Hospital, London, UK
Here we present a novel software package that allows easy creation and cross referencing of user databases of clinical spectra. The user-friendly GUI provides access to powerful statistical methods including Principal Components Analysis, Independent Component Analysis, and Linear Discriminant Analysis.
1621. Web-Based Graphical User Interface for the GE MRS/MRSI Data with LCModel
Cheng-Yun Hsu1, Cheng-Wen Ko1, Wei-Der Lui1, Martin Buechert2, Shang-Yueh Tsai3
1National Sun Yat-sen University, Kaohsiung, Taiwan; 2University Hospital Freiburg, Freiburg, Germany; 3National Taiwan University, Taipei, Taiwan
In this work we implemented an user interface based on web browser for LCModel access within an institution. It provides users a convenient approach to access LCModel and benefits the plug-in development due to the popularity of web scripts, e.g. PHP, Java, and HTML. We believe that this tool will be very helpful and valuable for MRS studies.
Gabriela Hossu1, 2, Francois Kauffmann3, Guillaume Calmon2, Sebastien Davard2, Muriel Perrin2, Sebastien Saez4, Christophe Dolabdjian4, Pierre Denise5, Andre Sesboue6, Jean-Marc Constans5
1CAEN University Hospital, CAEN, France; 2GE Healthccare, BUC, France; 3University, CAEN, France; 4GREYC CNRS UMR 6072 ENSICAEN, CAEN, France; 5CAEN University Hospital, CAEN, France; 6University, CAEN, France
Developed by Barantin ERETIC method consists of adding and synchronizing a pseudo-FID and metabolite signals electronically produced to calibrate. We present a simpler method ERETIC-II not requiring any synchronization. The ERETIC signal was continuously sent and unprocessed 1H 1.5T PRESS short TE spectra were acquired in full temporal resolution without averaging and separately analyzed with SCI-MRS-LAB program. Intra-variability and inter-recording CV were respectively 0.5% and 0.2%, showing improvements variability compared to previous results. This ERETIC-II method decreases variability, avoid difficulties of synchronization, could be applied in CSI and contributes to quality control in order to improve variability quantification in pathologies.
1623.
An Automatic Time-Domain Algorithm for the Quantification of
in Vivo Non-Water-Suppressed MR
Spectroscopy: The Filter-Diagonalization
Method (FDM)
Jyh-Miin Lin1, Shang-Yueh Tsai1, Hua-Shan Liu1, 2, Hsiao-Wen Chung1, Chou-Ming Cheng3, Tzu-Chen Yeh3, Robert V. Mulkern4, Nan-Kuei Chen5
1Graduate Institute of Biomedical Electronics and Bioinformatics, National Taiwan University, Taipei, Taiwan; 2Department of Radiology, Tri-Service General Hospital and National Defense Medical Center, Taipei, Taiwan; 3Taipei Veterans General Hospital, Taipei, Taiwan; 4Department of Radiology, Children's Hospital and Harvard Medical School, Boston, Massachusetts, USA; 5Brain Imaging and Analysis Center, Duke University, Durham, North Carolina, USA
It has been shown that the metabolite signals can be quantified more accurately using the non-water-suppressed (NWS) MRS technique, in which the water signals are fitted and removed in post-processing, because (1) the metabolite signals are not distorted by water suppression pulses and (2) the un-suppressed water serves as an internal reference for metabolite quantification. Here we propose to use the Filter-Diagonalization Method (FDM), an algorithm originally developed for quantum dynamic computation, to reliably and automatically quantify the metabolite signals in NWS MRS data with a very economic computational cost. In comparison to the existing post-processing methods, FDM has a much better localization property and can be applied to analyze a selected spectral range. Results from our simulation, phantom, and in vivo MRS studies demonstrate that the proposed technique is superior to the existing NWS MRS data processing methods
1624. Refinement of Simulated Basis Set for LCModel Analysis
Ivan Tkac1
1Univervity of Minnesota, Minneapolis, Minnesota, USA
Simulated bases set for LCModel analysis was refined using phantom metabolite spectra and highly resolved in vivo spectra from the rat brain at 9.4 T. Metabolite quantification using the new simulated basis set is in excellent agreement with the quantification performed using the experimental basis set. In addition, CRLB were significantly reduced. This update is mainly important for the weakly represented metabolites, such as Asc, Glc and GSH.
Hélène Ratiney1, Adriana Bucur1, Cristina Cudalbu1, Sophie Cavassila1
1CNRS UMR 5220, Inserm U630, INSA-Lyon, Université Lyon 1, Villeurbanne, France
The MR spectroscopic macromolecular signal, usually considered as a nuisance contribution in the quantification of short echo time signals, might reveal some interest as a disease marker by itself. Enabling its objective quantification and deriving a model function for its description is the purpose of the proposed method. Inversion recovery macromolecular spectrum coming from unresolved proteins, lipids and residual water shows broad patterns close to a gaussian mixture model. The proposed method takes advantage of the Expectation Maximization (EM) algorithm usually applied in machine learning or pattern recognition to provide a novel fitting procedure for broad patterns MR spectroscopic signal.
1626. Lineshape Accommodation in Quantitation of Magnetic Resonance Spectroscopy SignalsEmil Popa1, Hérald Rabeson1, Dirk van Ormondt2, Danielle Graveron-Demilly1
1Université Lyon 1, Villeurbanne, France; 2Delft University of Technology, Delft, Netherlands
Lineshape distortions due to residual eddy currents and magnetic field inhomogeneities are often present in short echo-time 1H spectroscopic data. If left uncorrected, these lineshape distortions lead to errors in metabolite concentration estimates when using quantification methods that incorporate model functions with specific lineshapes. In this study a new method is investigated, namely, the lineshape of the simulated metabolite basis-set signals is given the estimated lineshape of a reference spectrum before the quantitation step. Analytical formulae for the Cramér-Rao lower bounds on model function parameters of a Lorentzian and Gaussian singlet are also derived.
Novel Molecular & Cellular Imaging Agents
Hall D Tuesday 13:30-15:30
Cong Li1, Paul Winnard Jr1, Tomoyo Takagi1, Dmitri Artemov1, Zaver M. Bhujwalla1
1Johns Hopkins Univ., Baltimore, Maryland, USA
We have previously demonstrated the feasibility of performing image-guided prodrug enzyme therapy by incorporating MR and optical reporters on a prototype agent containing bacterial cytosine deaminase (CD), the enzyme that converts a non-toxic prodrug 5-fluorocytosine (5-FC) to 5-fluorouracil (5-FU). In a separate study we observed that siRNA-mediated downregulation of choline kinase increased the cell kill effects of 5-FU in breast cancer cells, but not nonmalignant breast cells. Here we are developing a strategy to incorporate siRNA within this prototype agent to generate 5-FU through conversion by CD within the tumor together with downregulation of choline kinase
1659.
Bacterial Gene Provides Cellular Contrast for MRI
Donna Elizabeth Goldhawk1, Claude Lemaire2, Savita Dhanvantari1, R. Terry Thompson1, Frank S. Prato1
1LHRI, London, Canada; 2University of Waterloo, Waterloo, Canada
Magnetotactic bacteria derive their magnetic properties from magnetosomes: membrane-bound, intracellular structures that form iron biominerals in response to the coordinated activity of approximately 20 genes. Similar to SPIO particles in size and composition, magnetosomes respond comparably to magnetic fields. We have investigated the ability of mammalian cells to produce magnetosome-like particles by expressing one of the bacterial genes: MagA, which encodes a putative iron transporter. High field MRI demonstrated that overexpression of GFP-MagA fusion protein increases cellular contrast in mouse neuroblastoma cells. These findings suggest that reporter gene expression systems for molecular MRI may be developed using magnetosome-related genes.
1660. A Gene Reporter System for Detection of Cellular LacZ Expression by Magnetic Resonance Imaging
Niclas Emanuel Bengtsson1, Glenn A. Walter1, Edward W. Scott1
1University of Florida, Gainesville, Florida, USA
The commercially available substrate S-gal reacts with â-galactosidase to produce a dark iron-rich precipitate that can be used as a genetic reporter system for simultaneous detection by MRI and histology. We found that S-gal labeling reduced T2* relaxation time for â-gal expressing bone marrow cells (BMCs) significantly more than control cells and the change in T2* increased dramatically with increasing magnetic field strengths. This decrease in T2* relaxation time translated into increased sensitivity and detection capabilities of transplanted â-Gal+ BMCs in vivo at 11.1T compared to 4.7T.
1661. Nanoprobes for 1H MRI Based Oximetry
Vikram D. Kodibagkar1, Xianghui Wang, Praveen Gulaka, Ralph P. Mason2
1UT Southwestern Medical Center at Dallas, Dallas , USA; 2UT Southwestern Medical Center at Dallas, Dallas, USA
There is increasing evidence for the importance of tissue oxygenation in development, progression, and response to cancer therapy. We present here synthesis, characterization and application of HMDSO based nanoemulsions suitable for intravenous delivery for 1H MR oximetry applications. The nanoprobes were prepared by ultrasonic emulsification of a mixture of HMDSO, Solutol® as surfactant and de-ionised water. Samples were bubbled with different oxygen mixtures and R1 (=1/T1) was measured as a function of pO2. The calibration curve for HMDSO based nanoprobes matched that of neat HMDSO. The nanoprobes show promise for in vivo measurements of pO2 and studies are currently underway.
1662.
Carbon Coated Microshells Containing Nanosized Gd(III)-Oxidic
Phases for Multiple Bio-Medical
Applications
Eliana Gianolio1, Aldo Arrais2, Stefano Avedano2, Giovanni Battista Giovenzana3, Enrico Boccaleri2, Mauro Botta2, Pier Luigi Staghellini2, Silvio Aime1
1Università di Torino, Torino, Italy; 2Università del Piemonte Orientale A. Avogadro, Alessandria, Italy; 3Università del Piemonte Orientale A. Avogadro, Novara, Italy
Herein is reported the high-yield synthesis and characterization of microsized graphenic carbon shells embedding nanometric Gd(III) oxidic phase to be applied as T2-susceptibility agents in MRI, X-ray scattering material in CT and activable substrates for Neutron Capture Therapy. The materials have been simply achieved by thermal treatment of raw humic acid precursor and GdCl3. They have been characterized by mean of multiple FT-IR and Raman, TGA, XRPD, XPS, SEM-EDAX and TEM techniques. The obtained capsular systems are featured with enhanced thermal and chemical stability and inherteness.
Cyrille Richard1, Bich-Thuy Doan2, Jean-Claude Beloeil3, Michel Bessodes4, Eva Toth3, Daniel Scherman1
1CNRS, Paris, France; 2CNRS, Gif sur Yvette, France; 3CNRS, Orléans, France; 4INSERM, Paris, France
A new T1 and T2 contrast agent has been developed, based on an amphiphilic gadolinium chelate adsorbed on multiwalled carbon nanotubes. The stable suspensions have been characterized with regard to MRI contrast agent applications. Relaxivities, r1, (20, 300 and 500 MHz) show a strong dependence on the GdL concentration, particularly at low field. Relaxation times T2 are practically independent of both the frequency and the GdL concentration. An in vivo feasibility MRI study performed at 300 MHz in mice revealed a negative contrast after injection of a suspension of functionalized nanotubes into the leg muscle.
1664. In Vitro Relaxivities Studies of Gadolinium Carbon Nanotubes at 0.2T
Kelvin Wong1, Jeyarama S. Ananta2, Samuel Patz3, Iga Muradyan3, Lon J. Wilson2
1The Methodist Hospital Research Institute, Houston, USA; 2Rice University, Houston, USA; 3Brigham and Women's Hospital, Boston, Massachusetts, USA
In this study, we characterize the r1 and r2 relaxivities of Gadonanotubes with different surfactant coatings at 0.2T using phantom imaging. The r1 relaxivity of bundled and individual Gadonanotubes were around 380 and dextran-coated Gadonanotubes has a high r1 at 800 mM-1s-1. All Gadonanotubes we tested have higher r1 relaxivities than previously reported NMRD studies. The r2 relaxivities of all Gadonanotubes are similar in the range of 300 mM-1s-1. Gadonanotubes is a strong positive MR contrast agents at 0.2T.
1665. In Vitro Relaxivities Studies of Gadolinium Carbon Nanotubes at 3T
Kelvin Wong1, Jeyarama S. Ananta2, Stephen Lin1, Lon J. Wilson2
1The Methodist Hospital Research Institute, Houston, USA; 2Rice University, Houston, USA
Gadolinium loaded ultra-short single wall carbon nanotubes have been recently proposed to be a high r1 relaxivity superparamagnetic MR contrast agent. Three Gadonanotube samples with different surfactant coatings were prepared and used in an in vitro phantom study. They were compared to two other contrast agents, a commercial gadolinium chelate and monocrystalline iron-oxide (MION). Our results show that bundled Gadonanotubes have at least three times higher r2* relaxivity compared to individual Gadonanotubes and about ten times higher r2* relaxivity than MION, indicating that Gadonanotubes may be a promising agent for in vivo cell labeling MRI.
Martin Lepage1, Jinqiang Jiang1, Jérôme Babin1, Bo Qi1, Luc Tremblay1, Yue Zhao1
1Université de Sherbrooke, Sherbrooke, Canada
A contrast agent was encapsulated into a photo-controllable nanocarrier and subsequently released upon absorption of UV light. We describe an in vitro MRI assay that can evaluate the state of incorporation of a small Gd-based contrast agent.
Réjean Lebel1, Beata Jastrzebska1, Hélène Therriault1, Marie-Michèle Cournoyer1, J Oliver McIntyre2, Emanuel Escher, Witold Neugebauer, Benoit Paquette1, Martin Lepage1
1Faculté de médecine, Université de Sherbrooke, Sherbrooke, Canada; 2Vanderbilt University Medical Center, Nashville, Tennessee, USA
A novel MRI protease-modulated contrast agent (PCA) was developed to detect the activity of matrix metalloproteinase-2 (MMP-2) in vivo. The PCA incorporates a solubility switch : upon cleavage by MMP-2, it becomes hydrophobic (PCA2-switch). This leads to an accumulation of PCA2-switch in an MMP-2-positive, MC7-L1 mammary carcinoma tumor in a Balb/c mouse model compared to a MC7-L1 MMP-2-knockdown tumor. When a scrambled peptide sequence is inserted into the PCA (PCA2-scrambled), the in vitro cleavage efficiency by MMP-2 is reduced and no accumulation is detected in vivo. In conclusion, PCA2-switch specifically accumulates in MMP-2-positive tumors.
1668. A New Bioactivated MRI Contrast Agent : Synthesis, Characterization and MR Imaging Studies
Yun-Ming Wang1, Yu-Zheng Su1, Tian-Lu Cheng1, Yu-Tong Kuo2, Twei-Shiun Jaw2, Gin-Chung Liu2
1Kaohsiung Medical University, Kaohsiung, Taiwan; 2Kaohsiung Medical University Hospital, Kaohsiung, Taiwan
A smart contrast agent, the β -glucopyronuronate-containing gadolinium(III) complex GdL1 (L1 = 1-(2-difluoromethyl-4-(1-(4,7,10-triscarboxymethyl-(1,4,7,10-tetraazacyclodecyl))acetamido)phenyl β -D-glucopyronuronate) was synthesized and characterized. Relaxometric studies show that the T1 change percentage in the value of GdL1 decreases dramatically (51%) in the presence of β -glucuronidase β -G) and human serum albumin (HSA). A significant signal change percentage enhanced by MR images was observed for GdL1 solution in the presence of β -G and HSA. The MR images also shows high intensity enhancement in CT26(+β -G) with β-G gene expression but not for the CT26 β -G) without β -G gene expression.
1669. Polymersomes: A New Tool in the Armoury of CEST Agents
Silvio Aime1, Daniela Delli Castelli1, Nicoletta Minari2, Alberto Sanino1, Enzo Terreno1
1University of Turin, Torino, Italy; 2University of Turin, T, Italy
The purpose of this work is to test the potential of polymersomes as novel platform for paramagnetic MRI-CEST agents. Liposomes loaded with paramagnetic shift reagents have been already proposed for this scope, but for in vivo applications it is necessary to improve the stability of the nanovesicles that are quite avidly taken up by macrophages. It has been reported that polymersomes display a higher in vivo stability than liposomes. The main result of this work is that the CEST properties of polymersomes are very similar to liposomes, thus making the former promising candidates for in vivo detection of CEST contrast.
1670. Novel Type of 19F MRI Contrast Agent
Stefanie Cordula Sparka1, Anne Werner1, Thomas Kampf1, Thomas Christian Basse-Lüsebrink2, Daniel Haddad2, Wolfgang Rudolf Bauer3, Peter Michael Jakob1, Wolfdieter A. Schenk1
1Universität Würzburg, Würzburg, Germany; 2Research Center Magnetic Resonance Bavaria, Würzburg, Germany; 3Universtätsklinikum Würzburg, Würzburg, Germany
A new type of 19F MRI contrast agent based on NH2-1B-DTPA were synthesized. Furthermore, MR-experiments were performed to analyze and visualize the different multi resonant 19F contrast agents.
1671. Synthesis of a Blood Brain Barrier (BBB)-Permeable MR Imaging Probe
Mohanraja Kumar1, Zdravka Medarova1, Guangping Dai1, Anna Moore1
1Massachusetts General Hospital, Charlestown, USA
The discovery of novel carriers capable of transporting imaging or therapeutic agents across the blood brain barrier is an important goal of neurological research. Here we report on a synthesis of a Gd-labeled fatty acylated polyarginine peptides (MPAP-Gd) with unique membrane translocation capacity. It was tested in mice injected intravenously MPAP-Gd. On T1 weighted images, there was a distinctive enhancement of the ventricles after injection of MPAP-Gd. Quantitative T1 map analysis revealed a marked shortening of the brain T1 immediately after injection. The ICP analysis of brain homogenates revealed an increase Gd content compared to mice injected with Gd-DOTA.
1672. Functionalizing Poly(Ethyleneimine) DNA Carriers with Gd-DOTA Contrast
Alexander B. Pine1, 2, Min Suk Shim2, Young Kwon2
1Cleveland Clinic, Cleveland, Ohio, USA; 2Case Western Reserve University, Cleveland, Ohio, USA
Recently we developed a method for functionalizing the polymeric gene carrier with an MRI contrastagent. Our technology involves DNA-encapsulating poly(ethyleneimine) (PEI) nanoparticles, which were fabricated in such a way that a controlled amount of MRI contrast agent, Gd-DOTA, was conjugated with primary amine terminals of PEI, which were subsequently condensed into DNA/PEI polyplexes. The presence of chelated paramagnetic gadolinium ions made the nanoparticles amenable to noninvasive detection using MRI, thus enabling a new approach to studying trafficking, biodistribution and localization ofthe polyplexes.
The increasing application of pharmacokinetic models in combination with DCE-MRI makes it indispensable to analyze whether different models yield comparable results. Thus, we performed a comparison of two frequently used pharmacokinetic models developed by Tofts and Brix. The comparison based on the simulation model MMID4 combined with measured AIFs and physiological muscle tissue parameters found in literature. Both models showed a good sensitivity for changes in perfusion and plasma volume but a low sensitivity for permeability changes. It also indicated that the Tofts model is more robust for extreme values but only if a sufficiently high tissue perfusion exists.
Hall D Tuesday 13:30-15:30
Different MR Contrast Agents
Shuning Huang1, 2, Christina Liu2, Guangping Dai2, Young Ro Kim2, Bruce R. Rosen, 12
1MIT, Cambridge, Massachusetts, USA; 2MGH, Charlestown, Massachusetts, USA
Magnetic resonance microscopy (MRM) has been successfully used in anatomical studies of various mouse strains and transgenic/mutant mouse models. Our data now demonstrate that MR contrast agents (Gd-DTPA and MnCl2) that have different relaxation properties and tissue affinity provide different image contrast, which shows well correlation in microscopic contrast with detailed cytocarchitecture.
1714. Uniplanar MR Stage Microscopy: Proof-Of-Concept Imaging and Distortion Correction
Lin Zhao1, Andrey V. Demyanenko1, Yun Kee1, Julian M. Tyszka1
1California Institute of Technology, Pasadena, California , USA
Initial experiences using an MR stage microscope to image millimeter scale samples are presented. An efficient model-based volumetric distortion correction scheme is also described. The stage microscope design has particular advantages for imaging relatively sensitive living samples such as developing embryos and cultured tissue explants since a much large volume of fluid medium can be used when compared to conventional volume coils with comparable sensitivity.
1715.
MRI Measurements of the Morphology and Vasculature of the
Mouse Eye in Vivo in Control Animals
and
Models of Diabetic Retinopathy
Thomas Neuberger1, Sebastian Aussenhofer1, Nadine Smith1, Mary Kennett1, Thomas Gardner1, Alistair Barber1, Andrew Webb, 1,
1Penn State University, University Park, Pennsylvania, USA
To develop an in vivo model for studying ocular diseases in mouse models using MRI the morphology and the vasculature were studied in wild type as well as in retinopathy developing C57BL/6J Ins2 Akita mice. Single retinal layers were detected with nine times higher spatial resolution than achieved previously. A comparison between retinal thickness in wild type and retinopathic mice did not show statistically significant differences, unlike previous studies in rats. High resolution 3D time-of-flight measurements allowed major blood vessels of the eye in the inner retinal layer and in the outer eye to be visualized for the first time.
1716.
Quantitative Magnetization Transfer Imaging for Evaluating
the Tissue-Engineered Cartilage
Weiguo Li1, Liu Hong1, Guoquan Zhang1, Richard Magin1
1University of Illinois at Chicago, Chicago, USA
Tissue-engineered cartilage is a promising approach for cartilage regeneration and repair. In this study, measurements of the development of tissue-engineered cartilage with quantitative magnetization transfer imaging (QMTI), T1, T2, and ADC were conducted over a 3-week growth period. These results were correlated with subsequent biochemical analysis for GAG contents. Bound proton faction (BPF) and magnetization transfer rate (k) show a statistical increase after one week in the tissue culture. This change was highly correlated with the increase of GAG. This study demonstrates that the QMTI can be used to access the changes of the ECM during engineered cartilage development.
Cheryl R. McCreary1, 2, Viktor Skihar1, 2, V Wee Yong1, 2, J Ross Mitchell1, 2, Jeff F. Dunn1, 2
1Unversity of Calgary, Calgary, Canada; 2University of Calgary, Calgary, Canada
Local injection of lysolecithin into the cervical spinal cord of mice provides an animal model of demyelination and remyelination. Volumes are currently assessed with serial histology, which is very time-consuming, and the results are highly variable. The aim of this study was to measure the size and extent of lysolecithin lesions using high resolution 3D gradient echo imaging of the fixed mouse spinal cord to characterize the lesion over a 4 week period.
1718.
Microscopic Skin Imaging at 7T
Stefan Maderwald1, 2, Oliver Kraff1, 2, Jens M. Theysohn1, 2, Marc U. Schlamann, 12, Mark E. Ladd1, 2, Harald H. Quick1, 2, Susanne C. Ladd1, 2
1Erwin L. Hahn Institute for Magnetic Resonance Imaging, Essen, Germany; 2Department of Diagnostic and Interventional Radiology and Neuroradiology, Essen, Germany
The high SNR of 7T MRI in conjunction with a 10-cm-diameter single loop transmit/receive coil opens up new perspectives for noninvasive imaging and characterization of human skin. In-vivo MRI with high resolution (voxel volumes of 10-2mm3 and smaller) over a 10 cm FOV within short examination times are feasible and were successfully performed in ten healthy volunteers, resulting in microscopic images with excellent quality. Detailed anatomic display of normal human skin as well as birthmark lesions is provided by this imaging concept, which includes the combination of a surface TX/RX RF coil at 7T and a dedicated high-resolution imaging protocol.
1719.
Perfluoropolyethers in Magnetic Resonance Microscopy: Effect
on Quantitative Magnetic Resonance
Imaging Measures and Histological
Properties of Formalin-Fixed Brain Tissue
Harpreet Hyare1, 2, Caroline Powell1, John S. Thornton1, Harry Parkes1, Laura Mancini1, Tarek Yousry1, Sebastian Brandner1, Po-Wah So2
1Institute of Neurology, London, UK; 2MRC Clinical Sciences Centre, Hammersmith Hospital, Imperial College London, London, UK
Purpose of this study was to determine whether perfluoropolyethers (PFPE) for ex vivo Magnetic Resonance Microscopy (MRM) alters MRI properties and/or the histological properties of fixed tissue. 4 of 8 formalin-fixed CD1 mouse brains were immersed into Fomblin® for 48 hours and 4 remained in 10% formol-saline as controls. After 2 days all were transferred to Fomblin® for MRI at 9.4T. No significant differences in any quantitative MR measure between the control and Fomblin-immersed mouse brains noted and no visible difference in anatomical or cellular detail noted, confirming suitability of PFPE as an embedding medium for MRM of fixed tissue.
1720.
Rapid MR Microscopy of Mouse Inner Ear Structures in Vivo
Using Linear Combination Steady-State
Free Precession MRI
H. Douglas Morris1, John A. Butman1
1National Insititutes of Health, Bethesda, Maryland, USA
MRI is an excellent method for visualizing the anatomical structures of the inner ear. Murine models of congenital deafness mutations can be quantified by microscopic MR imaging. Steady-state free precession imaging is highly selective for the endolymphatic fluid in the inner ear. Field distortions in SSFP images from nearby air filled structures can be reduced or cancelled by using a linear combination of SSFP images with different RF phase progressions. We demonstrate the utility of LCSSFP imaging in visualizing the inner ear structures of mice at 7.0T.
1721. Proton NMR and MRI Study of Sub-Millimeter Sized Biological Objects
Seongjin Choi1, 2, Peter Christopher Hammel1, David Tay1, Petra Schmalbrock1
1Ohio State University, Columbus, USA
NMR microscopy has been used for biological research but is particularly challenging for miniature objects. Although typical NMR experiments on small seeds suffered line-broadening, oil rich seeds could be distinguished from other seeds with standard liquid NMR studies. Conventional NMR and MRI are still useful for studying seed imbibition. In our study, water-uptake in seed material in an oxygen limited environment was observed. Thus MR microscopy can be used as a complementary method to conventional seed conductivity which assesses seed quality. With further development, the current NMR microscopy system could be used for MR biopsy.
1722. Noise Characterization and SNR Benefit of Cryogenic RF Coil
Xiaoli Zhang1, 2, Jun Dazai1, R Mark Henkelman1, 2
1Toronto Centre for Phenogenomics, Toronto, Canada; 2University of Toronto, Toronto, Canada
Signal-to-noise ratio is usually the major limitation to achieve higher spatial resolution in magnetic resonance microscopy such as mouse imaging. As the coil size decreases to a certain level, it becomes coil noise dominant. The purpose of this study was to understand coil and sample noise for surface coils as frequency and coil size change. A crossover point where coil noise equals sample noise at 7 T was found to be 2.1 cm in radius. SNR ratio of cryogenic coil to room temperature was predicted and compared with experimental results.
1723. Magnetic Resonance Microscopy of Blood Digestion in Ixodes Scapularis
Kevin J. Hallock1
1Boston University School of Medicine, Boston, Massachusetts, USA
Hematophagous arthropods infect millions of people every year with a wide variety of diseases including Lyme disease. Lyme disease is transmitted by several ticks, but the primary vector in the northeastern United States is Ixodes scapularis. Blood feeding and digestion are essential for I. scapularis development, but few methods exist to study these processes in vivo. Here we present the first serial magnetic resonance microscopy (MRM) investigation of blood digestion in I. scapularis and report the contrast changes caused by the digestion of hemoglobin. These results demonstrate that MRM is a promising tool for studying blood digestion in ticks.
1724. Enhancement, De-Enhancement and Contrast Locking of Fixed Rodent Brains
Hargun Sohi1, Daniele Procissi1, Xiaowei Zhang1, Julian Michael Tyszka1
1California Institute of Technology, Pasadena, California , USA
Contrast enhancement of fixed rodent brain tissue samples is an essential component of high SNR efficiency MR histology at high magnet fields [MR microscopy]. There is however very little information regarding the rate of ingress and egress of contrast agent from fixed tissue. We present estimates for the enhancement and de-enhancement time constants for perfusion fixed mouse brains within the skull, and propose a method for halting or locking the contrast enhancement at a desired stage for longer term storage prior to MR imaging.
Hall D Tuesday 13:30-15:30
Azzurra Grassi1, Leila Cammoun1, Claudio Pollo, Patric Hagmann2, Reto Meuli2, Jean Philippe Philippe Thiran1
1Signal Processing Institute, Ecole Polytechnique Federale de Lausanne, Ecublens, Switzerland; 2Centre Hospitalier Universitaire Vaudois and University of Lausanne, Switzerland
Thalamic nuclei can be distinguished by their characteristic fiber orientations, which influence the diffusion. Fiber orientations are relatively aligned within a nucleus due to the fact that the cerebrocortical striations within a nucleus all target the same region of cortex. The number of thalamic nuclei reported with histological methods varies with the method employed, although most cyto/myeloarchitectonic stains identify 14 major nuclei. We present a new approach for thalamic nuclei segmentation on High Angular Diffusion Resolution Images (HARDI), performed with a constrained k-means clustering.
1778.
Application of Discrminant Principal Component Analysis to
Distinguish Schizophrenic Subjects
from
Normal Controls Based on
Fractional Anisotropy Measurements
Arvind Caprihan1, Godfrey D. Pearlson2, 3, Vince D. Calhoun1, 4
1The MIND Institute, Albuquerque, USA; 2Olin Neuropsychiatry Research Center, Hartford, USA; 3Yale University, USA; 4University of New Mexico, Albuquerque, USA
In a pattern classification study the standard PCA can eliminate features that discriminate and keep features that best represent the group. We study an alternative method for selecting features which maximize the Mahalanobis distance between groups. We apply this discriminatory PCA (DPCA) to a fractional anisotropy data set from schizophrenic subjects and normal controls. We show that the choice of features can have significant impact if not done in a manner consistent with the final objective. For our data we were able to reduce the image dimensions from 149206x90 to a 60x90 matrix without any significant increase in classification error.
1779. Agreement and Disagreement Between Two Models of Diffusion MR Signal
Matt G. Hall1, Daniel C. Alexander1
1University College London, London, UK
We compare two models of diffusion in white matter undergoing inflammation: an analytical model and a monte-carlo model. We use both models to synthesise diffusion-weighted signals from spins diffusing in an environment undergoing cell swelling. We find good agreement in the models until the cylinders begin to abutt and overlap, when the monte-carlo model deviates from the analytical. We infer that cylinder abutting can have measurable effect on diffusion-weighted signal.
Matt G. Hall1, Daniel C. Alexander1
1University College London, London, UK
We compare approximations of cylinders of circular cross-section with cylinders with regular polygonal cross-section as restricting geometries in Monte-Carlo simulations of diffusion used to generate synthetic diffusion-weighted measurements. We find that a cylinder with six facets gice a good approximation of a circular cross section. We also present preliminary results approximating a sphere with platonic solids. We find that a tetrahedron has low anisotropy.
Tsang-Wei Tu1, Joong Hee Kim2, Hsiao-Fang Liang2, Sheng-Kwei Song2
1Aerospace and Structural Engineering, Washington University School of Engineering, St. Louis USA, Missouri, USA; 2Washington University School of Medicine, St. Louis, Missouri, USA
SynopsisA two-direction diffusion encoding gradient scheme to derive the directional diffusivities for assessing the extent of white matter damage in a mouse model of contusive spinal cord injury (SCI) has been examined and compared with the commonly used 6-direction DTI. Comparable results have been observed in diffusion anisotropy and directional diffusivities for the control mouse spinal cords. More importantly, the extent of white matter injury estimated using 6-direction DTI was comparable with those determined by 2-direction DWI. All in vivo MRI findings were validated with immunohistochemisty. The results suggest that the full tensor analysis is not necessary to quantity the extent of white matter injury in mouse SCI. This approach offers the advantage of saving a 67 % scan time needed for the conventional six-direction DTI.
1782. Parsimonious Model Selection for DTI Tissue Segmentation and Classification Applied to Clinical Data
Raisa Z. Freidlin1, 2, Evren Ozarslan1, Peter J. Basser1, Carlo Pierpaoli1
1NIH, Bethesda, USA; 2George Washington University, Washinton, District Of Columbia, USA
We used a statistical method for parsimonious model selection to choose the most appropriate water diffusion model (isotropic, oblate, prolate, or general anisotropic) for in vivo DTI of the human brain. We found that the prolate model reliably identifies structures with orientationally coherent fibers (e.g., the mid sections of the corpus callosum). However, most white matter regions are identified by the general anisotropic model, suggesting that their underlying fiber architecture is consistent with multiple fiber population within a voxel. The information provided by this method may be useful for optimal experimental design of diffusion MRI experiments and for selecting approaches to data analysis.
1783. The Elliptical Cone of Uncertainty in Diffusion Tensor Imaging and Its Normalized Measures
Cheng Guan Koay1, Peter J. Basser1
1National Institutes of Health, Bethesda, Maryland, USA
This study presents a technique of constructing the cone of uncertainty (COU) of the major eigenvector of the diffusion tensor based on the inverse of the Gnomonic projection, and two normalized geometric measures associated with the COU—the normalized areal and circumferential measures. These measures are local coherence measures for quantifying tract dispersion. The proposed measures, which are directly linked to the uncertainty in the major eigenvector of the diffusion tensor, may be important for probing the integrity of the white matter tracts in the brain and for DT-MRI tractography.
1784. Diffusivity in the Brain White Matter: Application of the Recursive Effective Medium Approximation
Oleg Petrovich Posnansky1, Nadim Jon Shah1
1Institute of Neurosciences and Biophysics (Medicine), Research Centre Juelich, Juelich, Germany
In order to explore the effects of a large range of geometrical and physical structural microparameters on the effective apparent diffusion coefficient we have developed a recursive effective medium approximation. The key point of this approximation consists in step-by-step averaging from the micro to the millimeter scale of the strongly fluctuating effective diffusivity. Such an approach provides insight into the fundamental of the diffusive process in a very random and heterogeneous structure of brain white matter and allows quantification of the sensitivity of the apparent diffusion coefficient to the variations of the different microparameters.
1785. Ellipsoidal Area Ratio (EAR): An Alternative Anisotropy Index for Diffusion Tensor Imaging
Dongrong Xu1, 2, Jiali Cui1, Peterson S. Bradley1, 2
1Columbia University, New York, New York, USA; 2New York State Psychiatric Institute, New York, New York, USA
Diffusion Anisotropy Index (DAI) is a scalar measure of diffusion tensor imaging data, and Fractional Anisotropy (FA) is the most popular one. However, FA is sensitive to noise and is not directly linked to the intuitive interpretation of a tensor¡¯s morphology. We have developed a novel DAI, called Ellipsoidal Area Ratio(EAR), that offers higher signal-to-noise ratios and maintain similar contrast-to-noise ratios.
1786. Dynamic Diffusion Analysis of Chitosan-Based Hydrogels Using Magnetic ResonanceRichard Wong1, Chih-Chang Chu, 12, Chao Zhong1, Yi Wang1, 3
1Cornell University, Ithaca, New York, USA; 2Cornell Unversity, Ithaca, New York, USA; 3Weill Medical College of Cornell University, New York, New York, USA
This study identifies three major metrics of hydrogel design and analyzes a chitosan-based family of hydrogels using MR techniques. Results demonstrate the behavioral variability that can be achieved by altering a hydrogel’s composition and illustrates to use of MR in synthesizing metric-based hydrogels.
Fiber Tracking & Connectivity Mapping
Hall D Tuesday 13:30-15:30
Kiran Kumar Seunarine1, Shahrum Nedjati-Gilani1, Matt G. Hall1, Philip A. Cook2, Daniel C. Alexander1
1University College London, London, UK; 2University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania, USA
Probabilistic tractography generates a map of connectivity of each voxel to a specified seed voxel, and typically use models of the uncertainty in each fibre-orientation estimate to provide fibre-direction samples for multiple repeats of the streamline tractography process. The assumption is that these uncertainty models reflect the true underlying distribution of fibre-orientations in each voxel, but may differ significantly in practice. We present an initial study into using the actual fibre-orientation distribution directly. Our method generates a calibration procedure to learn a mapping between features of the diffusion tensor and the fibre-orientation distribution and use that for probabilistic tractography.
1832. Combining Spherical Deconvolution and Streamline Tractography : Preliminary Results
Flavio Dell'Acqua1, 2, Paola Scifo1, Marco Catani2, Giuseppe Scotti1, Ferruccio Fazio1
1San Raffaele Scientific Institute, Milan, Italy; 2Institute of Psychiatry, London, UK
In this study, we combined a simple streamline tractography method with a modified damped Richardson-Lucy spherical deconvolution algorithm to perform dissections of white matter pathways in a complex brain region with several crossing fibers. Preliminary results on in-vivo data are shown.
1833. White Matter Tract Visualisation Using a Parabolic Eigensystem
Thomas Richard Barrick1, Nigel Lawes1, Chris A. Clark2
1Saint George's, University of London, London, UK; 2University College London, London, UK
Here we present a method for visual segmentation of tracts by interrogation of streamline termination coordinates and invocation of a parabolic eigensystem. This approach was designed to avoid problems associated with previous colour mapping schemes, in particular to avoid discontinuities in the colour mapping and reduction of computational time. The method is applied to a normalized brain dataset comprising 30 healthy individuals. Visualisations of brain stem structures, thalamic structures and the recently described three segment model of the arcuate fasciculus are demonstrated. Additionally, this technique provides several geometric parameters that may be useful for streamline segmentation algorithms.
1834. Whole Brain White Matter Tract Segmentation of Single Subject Diffusion Tensor Tractography DataThomas Richard Barrick1, Nigel Lawes1, Chris A. Clark2
1Saint George's, University of London, London, UK; 2University College London, London, UK
Here we present an algorithm for clustering of tracts throughout the white matter of the entire brain which requires the user to simply identify the constituent parts of the tract of interest. We show that this is robust when applied to DTI datasets of individual subjects. This allows rapid tract extraction to be performed without the need for group mapping or averaging of datasets, therefore maintaining characteristic features of individual subjects.
1835. Diffusion Tensor MRI Can Anatomically Segment Human Amygdaloid Subnuclei in Vivo
Eugenia Solano-Castiella1, Alfred Anwander1, Carol Docherty1, Enrico Reimer1, Marcel Weiss1, Angela D. Friederici1, Robert Turner1
1Max Planck Institute for Human Cognitive and Brain Sciences, Leipzig, Germany
The amygdaloid subnuclei are hard to distinguish using standard MRI. Their differential connectivity to other brain areas suggests that each might have a preferential internal fiber orientation. In this study we provide evidence for the segmentation of amygdaloid nuclei, using DTI data obtained in-vivo in 15 subjects and clustered with an automatic K-means algorithm. This finding may assist in discrimination of the distinct roles of these nuclei in processing of emotions and cognitive function, and in psychiatric disorders.
1836. Group Analysis of Human Brain White Matter Using Mean Path Analysis Method
Wen-Yang Chiang1, Van J. Wedeen2, Wen-Yih Isaac Tseng1
1National Taiwan University College of Medicine, Taipei, Taiwan; 2MGH Martinos Center for Biomedical Imaging, Charlestown, Massachusetts, USA
While globally optimized normalization was achieved, the performance of alignment for local structure, empirically, was not always guaranteed. In this study, we proposed a simple and effective strategy for group analysis of human white matter via co-registration of inter-subject mean paths of the tract bundles. Using the affine transform matrix estimated from coregistration of B0 images, the length of mean pathways of individual white matter could be normalized inhomogeneously. In addition to obtaining the averaged structural information, we can also analyze the change of the averaged tract bundle in a more reasonable way.
1837. A Ground Truth Analysis of the Preservation of Diffusion Tensor Information in a Population Specific Atlas
Wim Van Hecke1, 2, Alexander Leemans3, Steve De Backer1, Everhard Vandervliet2, Jan Sijbers1, Paul M. Parizel2
1VisionLab, University of Antwerp, Antwerp, Belgium; 2University Hospital Antwerp, University of Antwerp, Antwerp, Belgium; 3CUBRIC, University of Cardiff, Cardiff, UK
An important requisite for voxel based morphometry, studies is the use of a non-rigid coregistration technique, and the availability of a population specific atlas. In this context, aDTI atlas is created that preserves theorientational DT information and contains a minimal bias towards any individual image. To evaluate the proposed atlas, a ground truth method is developed using a single subject DT image that is deformed with 20 deformation fields. We demonstrate that the atlas contains all orientational diffusion information accurately and precisely, due to the use of all DT elements during coregistration and the specific atlas construction methodology.
1838. Probabilistic Tract-Based Atlas with High Angular Resolution Diffusion Imaging
Yi-Ping Chao1, Chun-Hung Yeh2, Kun-Hsien Chou2, Kuan-Hung Cho1, Jyh-Horng Chen1, Ching-Po Lin2
1National Taiwan University, Taipei, Taiwan; 2National Yang-Ming University, Taipei, Taiwan
In this study, we proposed a tract-based transformation to realign the multi-fiber trajectories with high angular resolution diffusion imaging. Transformation matrices were estimated by coregistering T1-weighted image of each subject to MNI-152 brain image using a 12-parameters affine linear registration. The matrices were then applied to the propagated neural bundles from individual subjects derived from QBI with MFACT tracking algorithm. By summarizing individual transferred tracts in the standard MNI-152 coordinate, a probabilistic tract-based atlas can be generated to assist clinical studies into the variations of complex connectivity between different subjects.
1839. A Comparison of Seeding Strategies for Group Tractography
Jonathan Daniel Clayden1, Amos J. Storkey2, Mark E. Bastin2
1University College London, London, UK; 2University of Edinburgh, Edinburgh, UK
The aim of this work was to illustrate and discuss the effects of applying common seeding strategies for tractography to a typical data set. Single and multiple seed approaches are compared with regard to their usefulness for robust segmentation in groups. We also examine the effects of thresholding on the results. We find that single seed strategies can be successful, whilst ROI methods using a group of seed points may not be as reliable as is often assumed.
Jonathan Daniel Clayden1, Amos J. Storkey2, Mark E. Bastin2
1University College London, London, UK; 2University of Edinburgh, Edinburgh, UK
This work describes how a probability distribution based on topological tract matching can be used to combine information derived from tractography for multiple seed points. This process provides broader coverage of a white matter structure of interest than data from a single seed voxel would, whilst maintaining high confidence in the segmentation. We also demonstrate that this approach mitigates the tendency for voxels with very high probability of connection to the seed point to be overrepresented in tractographic output.
1841. Application of a Double Inversion-Recovery Sequence to Diffusion Tractography
Simon J. P. Meara1, David M. Morris1, Karl V. Embleton1, Hamied A. Haroon1, Geoffrey J. M. Parker1
1University of Manchester, Manchester, UK
Areas of activation detected by functional magnetic resonance imaging (fMRI) experiments have previously been used to define seed regions for diffusion tractography. The present study has used a white matter-selective double inversion-recovery image to restrict regions of interest (ROIs) derived from fMRI results to contain only grey matter. It was found that the tractography results showed loci where a higher probability of connection was detected when using the whole ROIs, as compared to using the restricted ROIs. It must therefore be ensured for future studies that the ROIs used do not contain any of the surrounding white matter.
Bram Stieltjes1, 2, Jan Klein3, Ben Hyman2, L G. Naul2, Val Runge2, Marco Essig1
1DKFZ, Heidelberg, Germany; 2Scott and White Hospital, Temple, USA; 3MeVis Research, Bremen, Germany
The combination of parallel imaging (PI) and high field strength holds great promise for the improvement of EPI-based DTI and fiber tracking as shown previously. Here we show the first result using an experimental 32 channel head coil for depiction and quantification of small, strongly curved fibers such as the fornix and the cingulum. We show that within 4 minutes data can be obtained that allow for a reproducible quantification of fiber integrity in these structures. This method can be used for investigation of these structures in psychiatric diseases like Alzheimer’s and autism.
1843.
Acquiring Optimal DWI Data for Tractography on Post Mortem
Brain Tissue
Tim Bjørn Dyrby1, Jacob Jelsing2, Daniel C. Alexander3, Lise V. Søgaard1
1Copenhagen University Hospital Hvidovre, Hvidovre, Denmark; 2Copenhagen University Hospital Bispebjerg, Bispebjerg, Denmark; 3University College London, London, UK
The use of fixated post-mortem brains allows validation of mathematical models in a neuronal environment close to in vivo. However, lowered temperature and the process of fixation decrease the diffusion coefficient observed post-mortem. A wide range of b-values were investigated on a perfusion fixated pig brain. DWI data was acquired in one continuous scanning session and reconstructed using two methods with the ability to detect single and complex fibre compositions. For detection of crossing fibers a lower bound of b>2475s/mm2 was found. An upper bound was found to be b<8181s/mm2, and likely dependent upon the noise floor.
1844. Visualization of the Cingulum Bundles In-Vivo Using Optimized MR Diffusion Spectrum Imaging
marzieh Nezamzadeh1, Van Wedeen, Ruopeng Wang, Yu Zhang1, Wang Zhan1, Karl Young1, Michael Weiner1, Norbert Schuff1
1Radiology, University of California San Francisco, san francisco, California , USA
Previous DTI studies have shown that diffusion measurements at the tract level of cingulum fibers might yield a powerful marker of incipient AD pathology. However, DTI results can be ambiguous, especially in areas of crossing nerve fibers. Diffusion spectrum imaging (DSI) is more powerful than DTI and likely more sensitive to alterations of the cingulum bundles in AD. Our aim in this in-vivo 4T study on humans was therefore to optimize DSI parameters, especially spatial resolution and diffusion encoding to be able to accurately resolve the cingulum bundles for tract level measurements of diffusion. This study significantly demonstrates a reproducible identification of the cingulum bundles of human brain based on DSI acquisition with only 12 minutes duration. The gold standard setting provided from this study will be used for detection of the cingulum fibers in aging, mild cognitive impairment and Alzheimer’s disease.
1845. Reconstruction of Brainstem White-Matter Fiber Systems with Combinatorial Tracking
Shlomi Lifshits1, Arie Tamir, Yaniv Assaf
1Tel-Aviv University, Tel-Aviv, Israel
Tractography enables delineation of white matter trajectories. Recently we suggested the combinatorial tracking framework in which the DTI data is mapped to a graph, were the center of each voxel is connected to its neighbor voxel centers by graph edges. Here, we applied combinatorial tracking to reconstruct two fiber systems in the Brainstem: the pyramidal decussation (PD) and the medial cerebellar peduncle (MCP). We compare the results of the two reconstruction algorithms: the most probable random walk path and reaction path algorithms. Although both successfully reconstruct the PD, only the later reconstructs the MCP in an anatomically consistent manner.
1846. Fuzzy Anatomical Connectedness Using Diffusion MRI: An Approach to Tractography of the Brain
Stamatios N. Sotiropoulos1, Christopher R. Tench1, Li Bai1
1University of Nottingham, Nottingham, UK
We present a new framework for tract reconstruction from diffusion-weighted MR images. Given that medical images are inherently fuzzy due to physical limitations of the imaging device, we utilise a fuzzy connectedness algorithm. The strength of connectivity between any two voxels is assessed by examining all possible pathways connecting them to find the strongest path. We show how this algorithm can be applied to tract extraction using local fibre orientation estimates. Maps of the calculated connectedness values can be used to infer connectivity between a seed voxel and all other voxels in the brain.
Arnaud Messé1, Vincent Perlbarg1, Guillaume Marrelec1, Mélanie Pélégrini-Issac1, Habib Benali1, 2
1Inserm and Université Pierre et Marie Curie, Faculté de médecine Pitié-Salpêtrière, Paris, France, Metropolitan; 2Université de montréal, Montréal, Canada
This work investigates anatomo-functional connectivity in the human brain. Functional networks were identified from resting-state fMRI data using Independent Component Analysis. Using DTI data from the same subjects, a probabilistic fiber tracking method revealed the anatomical structure of the networks. Functional and anatomical connectivity indices were defined (functional correlation and mean number of fibers between pairs of regions in all networks, respectively) and a Multidimensional Scaling analysis was used to study the functional and anatomical organization of the networks. Results show similarities between functional and anatomical connectivities, supporting the existence of an underlying anatomical substrate for the functional networks.
1848. Fast-Marching Tractography for Connection Matrix (Fast-TraC)
Xavier Gigandet1, Leila Cammoun1, Reto Meuli2, Jean-Philippe Thiran1, Patric Hagmann1, 2
1Ecole Polytechnique Fédérale de Lausanne, Lausanne, Switzerland; 2Centre Hospitalier Universitaire Vaudois and University of Lausanne, Lausanne, Switzerland
Although high angular resolution diffusion MRI techniques are able to solve multiple intra-voxel fiber orientations, the usual streamline Diffusion Spectrum Imaging (DSI) tractography algorithms present some limitations in their ability to map complex fiber-crossings in the brain white matter because they select locally only the most linear trajectories. In this work, we present a fast marching tractography algorithm for DSI, called Fast-TraC, which 1) is able to efficiently address this issue, 2) creates fiber trajectories between 1000 small cortical ROIs covering the entire brain and 3) builds a whole brain connection matrix. We also see selected tracts that are accurately reconstructed.
1849. The Application of Point Spread Function Weighting in Probabilistic Tractography
David Mark Morris1, Sha Zhao1, Geoff JM Parker1
1University of Manchester, Manchester, UK
In probabilistic tractography linear interpolation is generally used to determine the weighting of the probability of selecting the streamline prorogation direction from neighbouring Probability Density Functions. Using the point spread function of the imaging acquisition has the potential to more realistically determine the weightings and improve the probabilistic tracking results. These point spread function weighting were determined the probability of selecting a particular Probability Density Function is compared between this implementation and the linear weighting at different positions. Differences observed are variable potentially influencing tractography results depending on the tracking environment.
1850. DTI Fiber Volume Measurement Using a Wrapping Algorithm
Bin Chen1, 2, Allen W. Song1
1Duke University, Durham, North Carolina, USA
DTI fiber tractography can potentially lead to quantitative assessment of brain development and pathology, provided that the size and volume of the neuronal fibers can be accurately measured. However, the relatively low resolution of DTI makes the voxel counting technique, the most commonly used method to estimate the fiber volume from tracked trajectories, less accurate because it has inherent partial volume effect. A fiber wrapping algorithm is proposed here to reduce such an effect for more accurate volume measurement.
1851. Development of DT-MRI Muscle Fiber Tracking Algorithms
Bruce M. Damon1, Zhaohua Ding1
1Vanderbilt University, Nashville, Tennessee, USA
Diffusion-tensor MRI is a promising tool for structural characterization of muscle, including fiber tracking. Partial volume artifacts due to intramuscular fat deposition require the development of new tracking algorithms capable of proper muscle fiber trajectories on the basis of readily measurable MR parameters. Here we use Monte Carlo simulations to test two sets of stop criteria and four interpolation schemes. We found that stop criteria using linear weighting of T2-weighted signal, FA, and curvature combined with nearest-neighbors interpolation provided optimum tracking results.
1852. Comparison of DT Tractography Algorithms with MEMRI and BOOT-TRAC
Chia-Ling Chen1, Kuan-Hung Cho2, Ke-Hsin Chen1, Ching-Po Lin1
1National Yang-Ming University, Taipei, Taiwan; 2National Taiwan University, Taipei, Taiwan
Three tractography algorithms, FACT, RungeKutta method (RK), and TEND, are evaluated by bootstrap tractography (BOOT-TRAC) and manganese-enhanced MRI (MEMRI). By changing the number of averages from 1 to 7, the success rate and the tolerance to noise are assessed. The results show that the higher success rate is obtained for all algorithms while higher SNR is achieved. For RK, the step size may also affect the results of tractography. In addition, the success rate of RK is generally higher than that of FACT and TEND. RK also gives a higher reliability for estimating neuronal connections.
1853.
Establishment of the In Vivo Tracing of Spinal Pathway
Using Diffusion Tensor Tractography
in Nonhuman Primates
Kanehiro Fujiyoshi1, Masaya Nakamura, Masayuki Yamada, Keigo Hikishima, Junichi Yamane, Hiroyuki Katoh, Kazuya Kitamura, Osahiko Tsuji, Seiji Okada, Akio Iwanami, Suketaka Momoshima, Yoshiaki Toyama, Hideyuki Okano
1Keio University school of medicine, Shinjuku, Japan
The evaluation of axonal fibers is essential to assess the spinal cord injury (SCI), but conventional methods are highly invasive, precluding clinical applications. We therefore sought to establish a non-invasive method to evaluate axonal fibers in vivo using diffusion tensor tractography (DTT). In this study, we performed DTT of both intact and injured spinal cords in common marmosets and confirmed the accuracy of DTT through histology. Our results revealed that DTT can depict the course and the disruption of specific neural pathways even in live animals, demonstrating the possible contribution of DTT to the clinical studies of SCI therapy.
1854. Simulations of DTI Voxel Size Resolution on Fiber Tract Measurements
Daniel T. Franc1, Kelvin O. Lim
1University of Minnesota, Minneapolis, USA
Diffusion tensor imaging (DTI) is being increasingly used to determine the orientation, size, and degree of anisotropy of individual white matter tracts for research studies and clinical practice protocols. The relatively low resolution for DTI leads to voxels that will partially sample anatomically separate tracts throughout the white matter when tracts have a similar size or smaller. This study uses modeled fiber tracts downsampled at different voxel resolutions to measure the uncertainty inherent in tract-specific DTI region-of-interest and fiber tracking measurements.
1855. A Comparative Study of Diffusion Tensor Field Transformations
Madhura Aditya Ingalhalikar1, Jinsuh Kim1, Vincent A. Magnotta1, Andrew Alexander2
1University of Iowa, Iowa City, Iowa, USA; 2University of Wisconsin, Madison, Wisconsin, USA
Diffusion tensor (DT) magnetic resonance imaging is unique in its ability to non-invasively visualize the white matter fiber tracts in human brain in vivo. MR diffusion is based on microscopic diffusion of water molecules which is a truly three dimensional process. A symmetric second rank tensor is often used as a model for characterizing the diffusion. Diffusion tensors contains orientational information that is affected by the transforms applied during image registration. Therefore, this directional information needs to be correctly rotated such that the tensors are consistent with the tissue reorientation caused by application of the resulting transformations. Previous work carried out in this area concentrates on reorientation of tensors by different methods. Xu et al (2003) have used the transformation and fiber direction estimate to reorient the tensors. Alexander et al (2001) have implemented the preservation of principle direction method. We have addressed this problem with a novel technique where the rotation in applied to the diffusion sensitizing gradients providing a voxel by voxel estimate of the diffusion gradients instead of a volume of by volume estimate. We have compared our technique with an existing method where the transformation is applied to the resulting diffusion tensors. In both the methods, rotation is computed from the deformation field by decomposing the local linear transformation.
Perfusion & Permeability: Dynamic Susceptibility Contrast Methods
Hall D Tuesday 13:30-15:30
Intravascular Signal?
Egbert Jan Willem Bleeker1, M. A. van Buchem1, M. J.P. van Osch1
1Leiden University Medical Center, Leiden, Netherlands
Local arterial input functions have less dispersion and delay than an AIF selected in the vicinity of the MCA. Another advantage is that flow-territory specific AIFs are used for deconvolution. It has been shown previously, that for local AIF measurements it is optimal to select voxels located completely outside the arteries. It could therefore be argued that crushing of vascular signal would improve local AIF measurements. The current study investigated the possible advantage of PRESTO, which crushes the vascular signal, over non crushing sequences at different echo times.
Sharon Chia-Ju Chen1, Keh-Shih Chuang1, Yuan-Yu Hsu2, Ho-Ling Liu3, 4
1National Tsing Hua University, HsinChu, Taiwan; 2Buddhist Tzu-Chi General Hospital-Taipei, Taipei, Taiwan; 3Chang-Gung University, Tao-Yuan, Taiwan; 4Chang-Gung Memorial Hospital, Tao-Yuan, Taiwan
Determination of arterial input function (AIF) is very critical in quantifying cerebral blood flow (CBF) in perfusion MRI study. Recent investigations suggest that the region of most T2* signal change during the passage of contrast agent is around vessels. Conventional region selection by hand drawing suffers from partial volume effect and cause error in the estimation of AIF. Independent component analysis (ICA) method decomposes signal into the interested source signal with alleviative partial volume effect. It provides more precise determination of AIF and quantification of CBF.
1897. DSC-MRI Perfusion Imaging: Which Input Function? – a Comparison to CT Perfusion Imaging
Irene K. Mikkelsen1, Doerthe Ziegelitz2, Birgitte F. Kjølby1, Göran Starck2, 3, Arnold Skimminge4, Maria T. Widmark2, Mats Tullberg3, Stig Holtås5, Carsten Wikkelsø3
1Århus University, Århus, Denmark; 2Sahlgrenska University Hospital, Göteborg, Sweden; 3Göteborg University, Göteborg, Sweden; 4Copenhagen University Hospital, Hvidovre, Denmark; 5Lund University Hospital, Lund, Sweden
The choice of artery for input function (AIF) measurement in dynamic susceptibility contrast imaging (DSC-MRI) influences the cerebral blood flow estimates (CBF). The DSC-MRI has no simple correction scheme for partial volume effects, whereas this is easily corrected for with CT. This study investigates the correlation between DSC-MRI and CT perfusion values for a typical CT and MR input functions. It is demonstrated that the correlation between DSC-MRI and CT are moderate for absolute perfusion values whereas the correlation of the GM/WM CBF ratio between modalities follows the identity line.
1898.
Does Deconvolution in DSC-MRI Deliver Bolus-Shape Independent
Results, or Why Noise is the Limiting
Factor in DSC-MRI?
Matus Straka1, Rexford D. Newbould1, Gregory W. Albers1, Roland Bammer1
1Stanford University, Stanford, California , USA
In quantitative DSC-MR PWI, deconvolution was aimed to reduce dependency of computed perfusion parameter on bolus shape. Theoretically, deconvolution should deliver perfusion parameter estimates independent of signal shape, but noise present in data and low-pass filtering induced by regularization causes dependency of CBF estimates on shape of the AIF, tissue curves and impulse response function of the underlying brain tissue. The error due to AIF shape change possible in typical measurements was found to be 10% of the true CBF value (20% of the estimates). Further, causes and solutions for the problem are discussed.
1899. What is a Good Sampling Rate for DSC-MRI Brain Perfusion Measurements?
Matus Straka1, Rexford D. Newbould1, Gregory W. Albers1, Roland Bammer1
1Stanford University, Stanford, California , USA
In quantitative DSC-MRI perfusion imaging, deconvolution process employs regularization to suppress noise-induced instabilities. Regularization therein has a strong low-pass filtering effect, hence removal of the high frequencies poses a question if a lower temporal resolution is acceptable as longer TRs are advantageous (less vessel signal clipping, more slices). Sampling at frequency related to the regularization cut-off leads to undesired changes due to spectrum aliasing, therefore sampling at higher frequency (related to spectra of signals and noise level) is proposed. Results confirm that sampling faster than necessary brings little benefit. For current MR PWI and noise, TR<2s might have no advantage.
1900.
On the Optimal Injection Speed for Bolus Tracking Perfusion
Peter Gall1, Irina Mader1, Birgitte Fuglsang Kjølby2, Valerij Kiselev1
1University Hospital Freiburg, Freiburg, Germany; 2Aarhus University Hospital, Aarhus, Denmark
The choice of an appropriate contrast agent injection speed and dosage is crucial for every DSC perfusion protocol. Contrast to noise (CNR) and temporal sampling are two competing measures for Gd-DTPA tracers. Simulations show that for typical bolus widths and sampling rates aliasing can occur. In order to avoid aliasing, CNR can be sacrificed by reducing the injection speed and therefore gain relative temporal resolution. The optimal injection setup was explored based on a study with 16 volunteers that underwent the measurement under different injection rates and dosages. The study shows that aliasing occurs for injection times smaller than 3*TR.
Kelvin Wong1, Chi-Pan Tam2, Michael Ng2, Stephen T.C. Wong1, Geoffrey Young3
1The Methodist Hospital Research Institute, Houston, USA; 2Hong Kong Baptist University, Hong Kong, People's Republic of China; 3Brigham and Women's Hospital, Harvard Medical School, Boston, Massachusetts, USA
In this study, we report the development of a model-independent deconvolution technique using least-absolute-deviation (LAD) regularization to improve the CBF estimation accuracy in the presence of bolus delay and bolus dispersion. Monte Carlo simulations were performed to compare the accuracy of CBF estimates derived from LAD, reformulated SVD (rSVD) and standard SVD (sSVD). The LAD method resulted in more accurate and reproducible residue function calculation than either rSVD or sSVD methods from no dispersion up to 1s dispersion.
1902.
A Novel Nonparametric Population Deconvolution for DSC-MRI
Quantification: Assessment on
Simulated Data
Denis Peruzzo1, Gianluigi Pillonetto1, Alessandra Bertoldo1, Claudio Cobelli1
1University of Padova, Padova, Italy
A novel nonparametric population deconvolution algorithm (PD), for DSC-MRI images quantification, is proposed and validated on simulated data. PD allows to simultaneously estimate the residue functions R(t) taking advantage of the entire collection of measures obtained from pixels population.PD provides more accurate estimates of CBF values than SVD, especially with dispersed R(t). R(t) obtained by PD are very regular, without unphysiological oscillations and closer to true R(t) than those obtained by SVD. When applied on a population having a bimodal CBF distribution, PD still provides closer estimates of R(t) than SVD, but less accurate CBF values.
1903. Model Free Bolus Arrival Time Estimation for Dynamic Contrast MR Studies
Arjan W. Simonetti1, Stefan Sunaert2, Ping Yang1, Arianne van Muiswinkel1
1Philips Medical Systems, Best, Netherlands; 2University Hospitals of the Catholic University, Leuven, Belgium
The estimation of bolus arrival time in dynamic MR data is relevant for clinical applications that rely on quantitative analysis like perfusion and permeability. We describe a method that uses the common lineshape of the MR signal intensity time curves and provides a sub-pixel correction of bolus arrival time. We show how this method can be used to select target pixels in arteries that may be used for arterial input function estimation and we show how relative bolus arrival time maps can be constructed.
1904. Comparison of an FT Based MMSE Method with OSVD Method for CBF Estimation in Patients with VCI
Unal Sakoglu1, Branko H. Garate1, Gary Rosenberg1, Rohit Sood1
1University of New Mexico, Albuquerque, New Mexico, USA
PWI using bolus-tracking technique based on DSC MRI has been developed to assess blood flow parameters such as CBF, CBV and MTT. Data processing of the acquired MRI data involves computation based on one of the two widely used convolution methods, circular SVD based method (oSVD) and the FT based method. In this study, a modified FT based MMSE (Minimum Mean Squared Error) method, that was proposed earlier and implemented as a simulation, has been used to extract flow parameters from PWI in 9 patients with Vascular Cognitive Impairment (VCI) and the results are compared with the existing standard oSVD method. The flow results are reported as relative CBF values, which is the ratio of CBF (normal white matter (WM) to gray matter (GM) and WM lesions (WML) to GM). Initial results demonstrate a significant decrease in relative CBF in the WM lesions (p<0.05, paired, two-tail t-test, n=9) with the results being more significant for the FT based MMSE method.
1905.
Using Dynamic Susceptibility Contrast MRI for Cerebral
Perfusion While Studying Alzheimer's Disease
Rachel DiAnne McKinsey1, Zhifei Wen2, Sterling Johnson, Alan McMcMillian1, Beth Meyerand2, Michele Fitzgerald, Cynthia Carlsson, Gemma Gliori, Sean B. Fain2
1University of Wisconsin Madison, Madison , USA; 2University of Wisconsin Madison, Madison, USA
The purpose of this study is to investigate the ability of DSC MRI with hybrid applications that combines bolus and infusion techniques to measure CBV and CBF changes in subjects with AD. Ten subjects with AD and ten Control Normal subjects were scanned on a GE Signa 1.5T MR. The statistical maps generated showed perfusion differences among AD and Control groups for CBV and CBF. Hypoperfusion was observed in both the Posterior Cingulate and thalamus for CBV and CBF. Other areas of decreased perfusion were caudate for CBV and posterior hippocampus for CBF. The data support DSC MRI as a viable means for measuring relative perfusion changes in AD.
1906. Modeling the Regulation of Cerebral Oxygen Extraction by Flow Heterogeneity
Sune Nørhøj Jespersen1, Mahmoud Ashkanian1, Kim Mouridsen1, Leif Ostergaard1
1Aarhus University, Aarhus, Denmark
The normal heterogeneity of blood flow velocities within the capillary bed has been speculated to constitute a reserve capacity, by which the oxygen extraction fraction can be increased through homogenization of microvascular flows, even at constant cerebral blood flow. For example, homogenization of the capillary flow distribution has previously been shown to be associated with high risk of subsequent infarction in acute ischemic stroke. Here we develop and analyze a comprehensive framework relating flow heterogeneity to the oxygen extraction fraction from dynamic susceptibility contrast perfusion raw data, and present an initial validation of the approach by comparison to O15 PET.
Jinsoo Uh1, Kelly Lewis-Amezcua1, Rani Varghese1, Hanzhang Lu1
1University of Texas Southwestern Medical Center, Dallas, USA
Vascular-Space-Occupancy (VASO) MRI is a new method for measuring cerebral blood volume (CBV) in humans, but several confounding factors need to be understood to fully optimize the method for accurate measurement. We performed VASO experiments (n=9) and simulations to investigate the effects of confounding factors including transverse relaxation, capillary water-exchange, and the change of concentration of contrast agent. The results demonstrate that water exchange can cause over-estimation, and T2/T2* effects can cause under-estimation in VASO CBV measurement. A post-contrast scan between 5-15 min with a multiple-echo acquisition scheme is recommended for accurate estimation of absolute CBV.
1908. Quantification of Cerebral Blood Volume Changes in Response to a Visual Stimulus in Humans
Nicholas P. Blockley1, Roman Wesolowski1, Susan T. Francis1, Penny A. Gowland1
1University of Nottingham, Nottingham, UK
It has previously been shown that dynamic changes in total cerebral blood volume (CBV), in response to a stimulus, can be monitored by using an infusion of contrast agent (CA). However this method only provides the percentage change in CBV. In this work we produce quantitative time-courses of CBV by integrating a DSC-MRI protocol into the existing paradigm. Rather than relying on a steadily increasing CA concentration provided by an infusion, here we rely on the exponential wash out of CA by the kidneys to measure the change in CBV. Pilot studies of this protocol were performed at 3.0 T and 7.0 T.
1909.
Steady-State Relationship Between Cerebral Blood Flow and
Venous Blood Volume
J. Jean Chen1, G. Bruce Pike1
1McGill University, Montreal, Canada
Grubb’s power coefficient has been widely used to derive total cerebral blood volume (CBV) from cerebral blood flow (CBF), but knowledge of the relationship between venous CBV (δCBVv) and CBF is scarce, yet it is well established that the BOLD signal is mainly modulated by venous CBV. In this work, we used the venous refocusing for volume estimation (VERVE) technique to measure δCBVv. We derived the steady-state relationship between δCBVv and δCBF in human subjects using graded visual and sensorimotor stimulation. We obtained a venous-specific power-law coefficient of 0.26, significantly lower than Grubb’s estimate of 0.38.
1910. High Resolution MR Vessel Size Imaging Using Dual Contrast Agent Injections
Wei-Shan Yang1, 2, Yuan-Yu Hsu3, Ho-Ling Liu1, 4
1Chang Gung University, Tao-Yuan, Taiwan; 2Cheng Hsin Rehabilitation Medical Center, Taipei, Taiwan; 3Buddhist Tzu Chi General Hospital, Taipei, Taiwan; 4Chang Gung Memorial Hospital, Tao-Yuan, Taiwan
Microvascular dimensions can be estimated by vessel size imaging (VSI) based on R2* and R2 measurements following bolus injection of contrast agents. VSI has been successfully applied in humans using a double-echo sequence which is however not widely available in clinical scanners and may be limited in spatial resolution. This study showed that VSI maps could be acquired with clinical sequences at 128x128 matrix during two separate dynamic contrast scans and the resulted vessel calibers agreed well with previously published data. Potential errors caused by temporal misalignment between two dynamic contrast scans were found within 8% based on computer simulation.
1911.
Comparison of Susceptibility-Weighted Imaging and Tracer
Kinetic Model Analysis in Brain Tumors
Chun-Jung Juan1, Hua-Shan Liu1, 2, Hing-Chiu Chang3, Hsiao-Wen Chung2, Chao-Ying Wang1, Queenie Chan4, Nai-Yu Cho5, Cheng-Yu Chen1, Chun-Jen Hsueh1, Chung-Ping Lo1, Guo-Shu Huang1
1Tri-Service General Hospital, Taipei, Taiwan; 2National Taiwan University, Taipei, Taiwan; 3GE Healthcare Taiwan, Taipei, Taiwan; 4Philips Medical Systems, Hong Kong; 5National Yang-Ming University, Taipei, Taiwan
This study was designed to evaluate the blood oxygenation level dependency (BOLD)-based vasculature and angiogenic status with microvascular proliferation in brain tumors by using susceptibility-weighting imaging (SWI) and the first-pass pharmacokinetic (FPPM) model of T2* MR perfusion-weighted images. We compared SWI and FPPM images with supplementary information from conventional contrast-enhanced T1-weighted (CET1) imaging to find associated pathological condition of vascular hyperplasia occurred on tumor proliferations. The preliminary results of observation show that CET1 excelled in tumor boundary detection. SWI excelled in visualizing blood vasculature and the FPPM model of perfusion-weighted imaging is good at detection in internal architecture than SWI and CET1.
Yi-Ying Wu1, 2, Ho-Ling Liu1, 2
1Chang Gung University, Kwei-Shan, Taiwan; 2Chang Gung Medical Hospital, Kwei-Shan, Taiwan
Disruption of BBB in brain tumors can cause errors in perfusion measurements using DSC MRI. Such errors were mainly caused by T1 effects in slightly leaky conditions and may be corrected using existing models. This study proposed a two-compartmental model and both T1 and T2 effects were taken into account in the signal time curve. In addition, the effects of the pre-loading dose were incorporated in the model. This model was showed to agree well with the existing ones in slightly leaky conditions and be able to simulate the additional T2 effects in very leaky conditions.
C. Chad Quarles1, Sunil Narayan1, John C. Gore1, Thomas E. Yankeelov1
1Vanderbilt University, Nashville, Tennessee, USA
DSC-MRI methods for assessing tumors are confounded by the extravasation of contrast agent resulting in simultaneous changes in the EES T1 and T2* that can confound susceptibility-induced signal decreases and yield unreliable CBV and CBF measurements. A methodological study of these relaxation interactions has yet to be reported. This is of note since the DSC parameter estimates could be significantly affected by both physiological and pulse sequence parameters. In this contribution we perform a series of simulations in which we systematically vary relevant MRI pulse sequence and physiological parameters and assess the errors returned in CBF and CBV.
1914.
Simultaneous Measurement of DSC- And DCE-MRI
Parameters Using Dual-Echo Spiral with a Standard
Dose of Gadolinium
Eric Scott Paulson1, Douglas Edward Prah1, Kathleen Marie Schmainda1
1Medical College of Wisconsin, Milwaukee, USA
We demonstrate here that DSC and DCE parameters, both corrected for confounding contrast agent effects, can be obtained simultaneously using a dual-echo spiral acquisition with a standard dose of gadolinium.
1915. Contrast Agent Extravasation Correction Combined with Automated AIF Identification in DSC-MRI
Qing Ji1, John O. Glass1, Wilburm E. Reddick1
1St.Jude Children's Research Hospital, Memphis, Tennessee, USA
We developed and implemented a simple and fast contrast agent leakage correction method in DSC perfusion MRI. In this method, the arterial input function (AIF) was first calculated from the perfusion MRI scan using a previously developed technique called iterative self-organizing map clustering technique. Then the AIF was directly used in a modified pharmacokinetic model to fit the concentration-time course in patient’s perfusion MRI scan. Because no extra parameters were introduced in the model, the fitting was fast. T2* weighted image data sets from fifteen pediatric brain tumor patients have been evaluated with this method. The results have demonstrated that both T1 and T2* effects caused by the contrast agent leakage can be corrected with this model.
MRA & Flow: Physiology & Diseases
Hall D Tuesday 13:30-15:30
Theodore J. Huppert1, Broc A. Burke, Solomon G. Diamond
1University of Pttsburgh, Pittsburgh, Pennsylvania, USA
Using high-speed (9.4Hz) T1-weighted functional MRI, we examine the physiological oscillations within the cerebral vasculature. We examine the temporal dynamics of signals from the major cerebral arteries and veins and explore the different characteristics of these signals and their relationships to systemic physiology.
1978. Improving Radial Sliding Window Cotrast-Enhanced Intracranial MRA Using HYPR
Hyun Jeong1, Christpher S. Eddleman1, Christopher Getch1, Matthew Walker1, Ty A. Cashen1, Timothy John Carroll1
1Northwestern University, Chicago, Illinois, USA
Intracranial contrast-enhanced angiography using radial sliding window reconstruction has been improved using HYPR processing. Using a sliding composite for the HYPR, significant increase in SNR was observed without significant changes in temporal profiles. Patients with angiographically confirmed AVMs were imaged.
1979. Hybrid STAR MR Angiography of the Intracranial Circulation
Robert R. Edelman1, 2, Ioannis Koktzoglou1, 2
1Evanston Northwestern Healthcare, Evanston, Illinois, USA; 2Feinberg School of Medicine, Northwestern University, Chicago, Illinois, USA
We propose a new approach to imaging of the intracranial circulation called hybrid STAR. Volunteers were imaged on a 1.5T Siemens Avanto system. Maximum lengths (in mm) for the ACA, MCA, PCA, and basilar artery (hybrid STAR vs. 3D TOF) were 113 ± 10 vs. 84 ± 17 (p < 0.01), 164 ± 10 vs. 126 ± 27 (p < 0.01), and 119 ± 11 vs. 89 ± 19 (p < 0.05). Hybrid STAR depicted significantly more branching vessels. Our preliminary results suggest the hybrid STAR technique provides much improved depiction of the intracranial circulation than has previously been feasible.
1980. Evidence of Vascular Steal in the Cerebral Circulation
Jeff A. Stainsby1, Julien Poublanc2, Adrian Crawley2, Daniel M. Mandell2, Jay S. Han3, Joe A. Fisher3, David J. Mikulis2
1GE Healthcare, Toronto, Canada; 2The Toronto Western Hospital, Toronto, Canada; 3Toronto General Hospital, Toronto, Canada
Measurements of Cerebral Vascular Reactivity using BOLD MRI provide can probe the autoregulatory capacity of the brain. Diseased tissue may often present with a negative reactivity (reduction in BOLD signal) following inhalation of carbon dioxide. This paper presents data comparing the CVR response, the timing of the response and the timing of blood flow arrival as measured by dynamic susceptibility contrast MR that supports the steal hypothesis used to explain these regions of negative reactivity.
1981. MR Susceptibility Weighted Imaging with Multi-Echo Acquisition
Yiping P. Du1, Zhaoyang Jin2, 3
1University of Colorado Denver School of Medicine, Denver, Colorado , USA; 2Hangzhou Dianzi University, Hangzhou, People's Republic of China; 3Zhejiang University, Hangzhou, People's Republic of China
A multi-echo 3D gradient-recalled echo pulse sequence was developed to improve the visibility of veins in susceptibility weighted imaging (SWI). Three echoes with echo times of 19.7ms, 23.4ms, and 37.6ms were acquired in a TR of 47ms. The veins were better depicted in short TE dataset in regions with severe field inghomogeneity due to reduced image artifacts. The veins in other regions have higher visibility in long TE dataset.
1982. High Resolution MR Susceptibility Weighted Imaging with Partial 3D K-Space Acquisition
Zhaoyang Jin1, 2, Ling Xia2, Yiping P. Du3
1Hangzhou Dianzi University, Hangzhou, People's Republic of China; 2Zhejiang University, Hangzhou, People's Republic of China; 3University of Colorado Denver School of Medicine, Denver, USA
Partial 3D k-space acquisition with partial echo and unsampled ky and kz views were used to reduce the scan time in susceptibility weighted imaging (SWI). Three-dimensional projection onto convex sets (POCS) algorithm was used for image reconstruction. Views at the corners of (ky,kz) domain were unsampled to further reduce scan time. A 3D Fermi filter was used to increase the signal-to-noise ratio and reduce angular dependence of spatial resolution. The effects of a 44% partial k-space acquisition and Fermi filter on the visibility of venous vasculature were demonstrated.
Michael Sibila1, 2, André Manuel Gaudnek1, 2, Monika Carola Lehmpfuhl3, Klaus Helmuth Obermayer1, 2, Andreas Hess, 24
1University of Technology, Berlin, Germany; 2Bernstein Centre for Computational Neuroscience, Berlin, Germany; 3Ludwig Maximilians University, Munich, Germany; 4Friedrich Alexander University Erlangen/Nuremburg, Erlangen, Germany
Since angiograms show blood vessels as spatially closely confined features, they can be used to register MR images of other types (e.g. fMRI) acquired in the same imaging session. This is only possible if MRA delivers reliable, reproducible images and does not show major random distortions. Therefore, we examine the reliability of MRA over subsequent scanning sessions and different imaging methods(2DTOF, 3DTOF,PC-MRA). Using an appropriate distance measure on our automatically generated geometric vasculature models, we examine the variance between different specimens in order to value the possibility of inter-specimen registration. We further use the rheological information contained in PC-MRA to investigate the reconstructed geometric models with respect to their function by means of fluid dynamics.
1984. Combination of Parallel Imaging and a Cut-Corner Acquisition for Neurovascular 4D-Flow
Thomas A. Hope1, Michael D. Hope, Roland Bammer, Marcus T. Alley
1Kaiser San Francisco, San Francisco, California , USA
Time-resolved three-dimensional phase-contrast MRI (4D-flow) is a tool for imaging blood flow velocity profiles, but is limited in its clinical applications by its long scan time. In this study, we analyze the effect of the combination of GRAPPA with an acquisition that removes the corners of k-space in order to further limit scan time. We also analyzed the effect on both mean velocity across the vessel lumen as well as the difference between individual velocity vectors, as individual vectors may effect calculations of wall shear stress and other variables that depend on the local velocity field. We show that the combination of a cut-corner acquisition to parallel imaging results in minimally decreased image quality when combined with GRAPPA while decreasing scan time by 20%.
1985.
Curve-Fitting Aided Pressure Gradient Assessment in Aqueduct
of Sylvius
Ming-Yen Chen1, Hsiao-Wen Chung1, Teng-Yi Huang2, Cheng-Yu Chen3
1National Taiwan University, Taipei, Taiwan; 2National Taiwan University of Science and Technology, Taipei, Taiwan; 3Tri-Service General Hospital and National Defense Medical Center, Taipei, Taiwan
Intracranial pressure (ICP) measurement was significant for diagnosis and treatment in hydrocephalus diseases. Noninvasive ICP MR estimation was proposed by Noam Alperin [1]. However, spatial and temporal limitations in cerebrospinal fluid (CSF) flow study of aqueduct of Sylvius, which were still bothered researchers¡¦ assessment. Since its small structure is deeply located in a brain, according to our past studies [2], the total pixels in aqueduct imaging were merely 9~25 in normal volunteers. Therefore, time-varying pressure gradient individually derived from each correspondingly spatial-averaged CSF velocity data, which might far identify with profile of velocities in these aqueduct-segmented pixels. This project aimed at evaluating through-plane flowing time-varying functions set by 2D curve-fitting method. By using a simplified Navier Stokes equation [1], we could depict deliberatively in ICP measurement. Here, we also compared different 2D curve-fitting methods, e.g. 4-order polynomial, parabolic and Gaussian fitting, when they applied to cine phase contrast (PC) MR images.
1986. Compare Image Segmentation Methods for Evaluating Aqueductal CSF Hemodynamic
Ming-Yen Chen1, 2, Hsiao-Wen Chung1, 3, Teng-Yi Huang4, Cheng-Yu Chen3
1National Taiwan University, Taipei, Taiwan; 2Yuanpei University of Science and Technology, Hsin-Tzu, Taiwan; 3Tri-Service General Hospital and National Defense Medical Center, Taipei, Taiwan; 4National Taiwan University of Science and Technology, Taipei, Taiwan
According to our past studies, even sacrificed outside of aqueduct region with partially spatial overlapped, the total pixels in aqueduct imaging were merely 9~25 in normal volunteers. Therefore, post-processing in image segmentation played an important role in calculating aqueductal area, stroke volume and related flow parameters. To avoid errors from inter-observer and intra-observer in region of interest (ROI) selection, we adopted automated image segmentation, including adaptive threshold in magnitude image, pulsatility based segmentation (PUBS) and independent component analysis (ICA). For accuracy and precision, both phantom and normal human study by using 2D cine magnetic resonance phase contrast imaging were compared.
1987. Venous Outflow of Label in ASL Perfusion MRI of Healthy Children and Children with Sickle Cell Disease
Wen-Chau Wu1, Hengyi Rao1, Mikolaj A. Pawlak1, Kim M. Cecil2, John VanMeter3, Thomas A. Zeffiro4, John A. Detre1, Elias R. Melhem1, Jiongjiong Wang1
1University of Pennsylvania, Philadelphia, Pennsylvania, USA; 2Cincinnati Children's Hospital, Cincinnati, Ohio, USA; 3Georgetown University Medical Center, Washington DC, USA; 4Massachusetts General Hospital, Boston, Massachusetts, USA
In perfusion studies of human brain using arterial spin labeling methods, it is generally assumed that all labeled water will exchange into brain tissue or relax before reaching the venous end. However, venous outflow of the label can present with elevated flow and prolonged blood T1 such as in a child’s brain. In the present study, venous outflow were detected as hyperintensities in the saggital sinus and further quantified as a function of the average flow in gray matter in healthy children and children diagnosed with sickle cell disease. Our data demonstrated significant venous outflow effects in the pediatric population.
Alan Jackson1, Deborah Sinclair2, Stavros Stivaros3
1University of Manchester, Withington, UK; 2Salford Royal Hospitals NHS Trust, Salford, UK; 3University of Manchester, Manchester, UK
We have examined the relationship between premorbid measurements of CSF and blood flow measurements to predict objective changes in brain volume and ventriculostomy flow in a group of patients undergoing third ventriculostomy for communicating hydrocephalus
1989.
Arterial Spin Labeling MRI Measurements of Timing Parameters
in Relation to Collateral Flow Patterns in
Patients with Carotid Artery
Occlusion
Reinoud Pieter Harmen Bokkers1, Peter Jan van Laar1, Kim C C van de Ven1, L J. Kappelle1, C J M Klijn1, Jeroen Hendrikse1
1University Medical Center Utrecht, Utrecht, Netherlands
Arterial spin labeling (ASL) at multiple delay times can be exploited in perfusion MRI imaging to visualize and quantify the temporal dynamics of blood inflow. This is the first study to investigate the consequences of internal carotid artery (ICA) occlusion and collateral blood flow patterns on regional timing parameters. Assessment of timing parameters in patients with ICA occlusion (n=17) demonstrated regional heterogeneity. With presence of leptomeningeal collaterals, impaired cerebral hemodynamics was found in the frontal region, indicating that ASL can be used to quantify the delay of arterial blood associated with collateral perfusion and the extent of such collateral perfusion.
1990.
Arterial Spin Labeling Perfusion MRI at Multiple Delay Times
in Patients with an Internal Carotid Artery
Occlusion: A Correlative
Study with 15O Positron Emission Tomography
Reinoud Pieter Harmen Bokkers1, Jeroen Hendrikse1, J P. Pluim1, J M. Bremmer1, C J M Klijn1
1University Medical Center Utrecht, Utrecht, Netherlands
Arterial spin labeling (ASL) perfusion MRI at multiple delay times has previously been introduced as an ASL method to compensate for the blood transit delays in patients with an internal carotid artery (ICA) occlusion. The aim of this work was to compare the use of ASL at multiple TIs and Oxygen-15 PET in patients with a symptomatic ICA occlusion. Assessment of CBF with both PET with ASL demonstrated a good correlation between CBF gray (r=0.54, p<0.01) and white-matter (r=0.53, p<0.01), indicating that ASL at multiple TIs can be used in clinical practice as a non-invasive technique to assess brain perfusion.
1991. Mid-Sagittal Saturated MRA for Assessment of Blood Flow from STA-MCA Bypass
Toshiaki Akashi1, 2, Toshiaki Taoka1, Hiroyuki Nakagawa1, Toshiteru Miyasaka1, Masahiko Sakamoto2, Satoru Kitano1, Satoru Iwasaki3, Kimihiko Kichikawa1
1Nara Medical University, Kashihara, Japan; 2Nara Prefectural Nara Hospital, Nara, Japan; 3Higashiosaka City General Hospital, Higashiosaka, Japan
Conventional MRA and contrast angiography have been selected as modalities to assess the bypass flow following STA-MCA bypass. However, conventional MRA can not differentiate whether the blood flow come retrograde from the bypass or antegrade. Whereas external carotid angiography (ECAG) can demonstrate the bypass flow, it is invasive and may not show natural flow because of pressure injection of contrast. We demonstrated usefulness of time of flight MRA with saturation band in the mid-sagittal area to suppress signal from internal carotid arteries (gMid-sagittal saturated MRAh) for visualization of natural bypass flow from STA-MCA, and agreement with the ECAG.
1992. Improvement in Patients Cerebrovascular Reserve Following Carotid Endarterectomy
Stephen Goode1, 2, Shane MacSweeney3, Dorothee Auer3
1Queens Medical Centre, Nottingham, UK; 2NUH, Nottingham, UK; 3NUH, UK
In patients with symptomatic carotid artery disease(CAD) the predominant mechanism causing ischaemic injury is considered to be thromboembolic, however cerebral haemodynamics can also be compromised. Currently data concerning haemodynamic status before and after carotid endarterectomy(CEA) is insufficient. The aim of this study was to assess the haemodynamic effects of CEA on CVR in patients with CAD using hypercapnia fMRI. 12 patients with symptomatic CAD were scanned. Following CEA there is a significant improvement in the ipsilateral CVR but no improvement in the contralateral territory. Assessment of the AI in these patients revealed that this asymmetry is actually reversed following intervention.
1993.
A Contrast Dose Reduction Study for 3D High-Spatial
Resolution Contrast-Enhanced Cerebral Magnetic
Resonance Venography at
3.0 Tesla
Anderanik Tomasian1, Noriko Salamon, Mayil Krishnam, J. Pablo Villablanca, J. Paul Finn
1UCLA, Los Angeles, USA
The purpose of our study was to prospectively establish the non-inferiority of diminished dose regimens compared to higher dose regimens as reflected in the diagnostic image quality of high-spatial-resolution three-dimensional MR angiography of the supra-aortic arteries at 3.0 Tesla.High-spatial-resolution cerebral CE-MRV performed comparably well with the two contrast dose regimens for evaluation of large intra-cranial venous structures at 3.0T
1994. 3D TOF MR Angiography at 3T vs. Catheter Angiography: A Quantitative Comparison
Sanjoy Nagaraja1, Kuan J. Lee1, David Capener1, Stuart Coley, Matthew Kaduthodil, Lee Walton, Jim M. Wild1, Andras Kemeny, Iain D. Wilkinson1, Paul D. Griffiths1
1University of Sheffield, Sheffield, UK
Arteriovenous malformations in the brain are complex vascular lesions that are a major cause of haemorrhagic stroke. We used MRA at 3T to measure nidal maximum linear dimensions, nidal volumes and the number of venous drainers, and compared the results with measurements made with the gold standard, catheter angiography.
1995. In-Vivo Plaque Imaging of the Carotid Arteries at 7 Tesla: First Results
Oliver Kraff1, 2, Stefan Maderwald1, 2, Steffen Hahn1, 2, Jens M. Theysohn1, 2, Mark E. Ladd1, 2, Joerg Barkhausen1, 2, Elke R. Gizewski1, 2, Harald H. Quick1, 2
1Erwin L. Hahn Institute for MRI, Essen, Germany; 2University Hospital Essen, Essen, Germany
Recently, imaging of the carotid arteries at 3T has demonstrated the advantages of high-field MRI in plaque depiction. To further push the limits of high-field carotid MRI, our study aimed to optimize different 2D and 3D GRE and TSE sequences for plaque imaging at 7T using a transmit/receive single loop coil. The sequences were first optimized in five healthy volunteers and subsequently tested in one patient. Plaque and degeneration of the vessel wall were clearly visible in the 7T images. SAR restrictions, difficulties with the ECG trigger, and high-field related image artifacts were identified as challenges of 7T carotid MRI.
1996.
3D High Resolution Susceptibility Weighted Imaging (SWI)
Venography at 3T and 7T
Yulin Ge1, Samuel Barnes, Samantha Heller1, Yingbiao Xu, Qun Chen1, E Mark Haacke, Robert I. Grossman1
1New York University School of Medicine, New York City, New York, USA
This work describes a markedly improved susceptibility weighted imaging (SWI) based venography that is acquired at 7T vs 3T. Preliminary results indicating the promise of using ultra-high field SWI to generate a high resolution and high quality venography by virtue of greatly increased signal-to-noise ratio (SNR) and susceptibility contrast.
Pediatric Normal Brain: Multimodality
Hall D Tuesday 13:30-15:30
Latha Srinivasan1, 2, Hui Xue1, Serena J. Counsell1, Joanna M. Allsop1, Julie A. Fitzpatrick1, Giuliana Durighel1, A David Edwards1, Mary A. Rutherford1, Daniel Rueckert1, Joseph V. Hajnal1
1Imperial College London, London, UK
Convolutions are the most striking feature of third trimester cortical development. 120 preterm infants were studied between 25 and 50 weeks gestation using a neonatal specific cortical segmentation and reconstruction technique. The cortical surface area expanded exponentially until 36 weeks after which the surface area and volume increased proportionally. Curvature analysis of the inner cortical surface revealed that the early surfeit in surface area corresponded with an increase in the intrinsic curvature whereas the later proportional growth mirrored an increase in the extrinsic curvature, suggesting that both intrinsic and extrinsic mechanisms are active albeit at different phases of cortical gyrification.
2045. MRI Children Atlas Based on a Robust Estimator
Juan José Ortiz1, Leopoldo José Gonzalez1, Roberto Emmanuele Mercadillo1, Fernando Alejandro Barrios1
1Universidad Nacional Autonoma de Mexico, Queretaro, Mexico
Brain atlases are used to project and localize activated structures in fMRI or statistical differences in BVM. Brain atlases or templates are averaged 3D image volumes projected into stereotactic space, generated from one or several subjects using different medical imaging techniques. Children templates pose interesting challenges since structure and spatial localization differs from other children during development and form the adult brain. In this work a one children template was generated using the mode as robust statistical estimator, and three different realignment processes are compared.
2046.
Age and Regional Dependent Changes of Glutamate in Human
Brain: In Vivo Quantitation with MR
Spectroscopy
Ashok Panigrahy1, Jane Tavare1, Marvin D. Nelson1, Floyd H. Gilles1, Istvan Seri1, Stefan Bluml1, 2
1Childrens Hospital Los Angeles/USC, Los Angeles, California , USA; 2Rudi Schulte Research Institute, Santa Barbara, California , USA
Age-dependent normal changes of glutamate concentrations were determined in 416 subjects with normal MRI and unremarkable clinical follow-up ranging in age from 25 weeks post-conceptional age (15 weeks premature) to 20 years. Glutamate concentrations increased from 3.7 ±2.0 mmol/kg at term to adult levels of 8.6±1.1mmol/kg in parietal white matter and from 4.4±1.5 mmol/kg to 12.9±1.3mmol/kg in occipital grey matter within the first year of life. The highest rate of net glutamate synthesis in WM was observed at 6 weeks after birth (0.46mmol/kg/week) and in grey matter 12 weeks after birth (0.36mmol/kg/week).
2047. T2 Relaxometry of Normal Pediatric Brain Development
Ilana Ruth Leppert1, C Robert Almli, Robert C. McKinstry, Robert V. Mulkern, Carlo Pierpaoli, Michael J. Rivkin, Gilbert Bruce Pike1
1Montreal Neurological Institute, Montreal, Canada
This work establishes normal age-related changes in MR T2 relaxation time constants using data collected as part of the NIH MRI Study of Normal Brain Development. The main finding, based on over 200 subject scans, is the parametrization of the mono-exponential evolution of T2 in several brain regions for children aged 0 to 4.5 years. This behavior is believed to reflect the rapid water content changes as well as myelination processes during neonatal brain development. These results represent a subset of a publicly available normative pediatric MRI database, providing a basis for comparison for studies assessing normal brain development and deviation due to various neurological disorders.
Jieun Kim1, In-Young Choi1, Yafeng Dong1, Carl P. Weiner1, Sang-Pil Lee1
1University of Kansas Medical Center, Kansas city, Kansas, USA
Guinea pigs are ideal for studying the growth and development of the brain due to their precocious neurological development. DTI provides good endogenous brain tissue delineation contrast. We obtained quantitative measurement for brain white matter development in postnatal guinea pigs using DTI in vivo. This work could pave a way to study abnormalities of brain development and to understand the underlying mechanisms of the most common cause of fetal brain damage, chronic fetal hypoxemia.
2049.
Comparative DTI Study of Preterm and Normal Newborns : SPM
and ROI Analyses
Paola Scifo1, 2, Cristina Baldoli1, 2, Silvia Pontesilli1, 2, Federica Navarra1, 2, Roberta Scotti1, 2, Antonella Poloniato1, Giuseppe Scotti1, 2, Ferruccio Fazio1
1Scientific Institute H San Raffaele, Milan, Italy; 2Università Vita-Salute San Raffaele, Milan, Italy
In this work, we show a comparative DTI study between a group of normal newborns and a group of preterm newborns these last acquired three times, at different stages of development. Analyses have been performed both on voxel-by-voxel basis and with ROIs. Results show a delayed myelination in preterm newborns with respect to normal subjects but this evidence is reduced along time. Comparison between SPM and ROI analyses show congruent results.
Deqiang Qiu1, Li-Hai Tan1, Pek-Lan Khong1
1The University of Hong Kong, Hong Kong, People's Republic of China
Using diffusion tensor MR imaging(DTI) and advanced voxelwise analysis tools, we aim to study changes of diffusivities and anisotropy of white matter in three age groups from late childhood to young adulthood (7.4yrs +/-0.3, n=24, 10.3yrs +/-0.5, n=27, 22.8yrs +/-2.3, n=24). We found predominantly increased FA and decreased MD with increasing age in regions of cerebellar white matter, right temporal white matter and a large portion of the superior frontal and parietal white matter driven by both the reduction of radial diffusivity and axial diffusivity but the latter to a lesser extent.
Catherine Lebel1, Christian Beaulieu1
1University of Alberta, Edmonton, Canada
The superior longitudinal fasciculus (SLF) is a white matter tract involved in many cognitive tasks. This study used diffusion tensor tractography to investigate SLF asymmetry in 223 healthy, right-handed subjects aged 5-58 years, and a subgroup of 59 children 5-13 years who also underwent cognitive assessment. We tracked the direct SLF segment in each hemisphere, and classified subjects as left-lateralized, symmetric, or right-lateralized. Most individuals were left-lateralized; however, some individuals, mainly children, showed rightward asymmetry. Symmetric children outperformed left-lateralized children on five of seven cognitive assessments and right-lateralized children on three tasks, suggesting symmetric arrangements offer a certain cognitive advantage.
2052. Relations Between White Matter Integrity and Working Memory in Children and Adolescents
Donald Mabbott1, 2, Joanne Rovet1, 2, Conrad Rockel1
1The Hospital for Sick Children, Toronto, Canada; 2University of Toronto, Toronto, Canada
In this diffusion tensor study we examined regional differences in the relations between age and hemispheric white matter fractional anisotropy (FA) and tested the relations between FA, age, and working memory in 36 typically developing children. FA was calculated for 12 regions of hemispheric white matter. The effect of age on FA was stronger for frontal-parietal than posterior white matter. Only right inferior-frontal regions contributed uniquely beyond the effect of age in accounting for working memory performance. Our findings are consistent with the growth of regional white matter organization as playing an important role in increased working memory with age.
2053.
Mapping of Paediatric Motor Function and Structure Using
Magnetoencephalography (MEG) and
Diffusion Tensor Imaging (DTI)
Donald Mabbott1, 2, William Gaetz1, Elysa Widjaja1, Sonya Bells1, Conrad Rockel1
1The Hospital for Sick Children, Toronto, Canada; 2University of Toronton, Toronto, Canada
MEG was used to identify motor fields for the left and right hands in young children and these activations were used as seeds for probabilistic tractography of the associated white matter pathways. Quantitative measures of the integrity of these tracts were calculated and we evaluated whether differences existed between the dominant and non-dominant hands, and age-related changes were evident. The Motor Field was localized to the hand area of pre-central gyrus with tractography delineating cortical-spinal tracts. FA for the dominant hand cortical-spinal tract was greater than for the non-dominant hand and age related changes were evident for the dominant hand only.
2054. Perfusion Imaging of Normal Brain Development from Childhood to Young Adulthood
Hengyi Rao1, Sumei Wang1, Joan Giannetta2, Kim M. Cecil3, John VanMeter4, Thomas A. Zeffiro5, Hallam Hurt2, John A. Detre1, Jiongjiong Wang1
1University of Pennsylvania, Philadelphia, Pennsylvania, USA; 2the Children's Hospital of Philadelphia, Philadelphia, Pennsylvania, USA; 3Cincinnati Children's Hospital, Cincinnati, Ohio, USA; 4Georgetown University Medical Center, Washington DC, USA; 5Massachusetts General Hospital, Boston, Massachusetts, USA
Continuous arterial spin labeling (CASL) perfusion MRI was used to examine the typical development of cerebral blood flow (CBF) in a cohort of 90 healthy pediatric (5-16yrs) and young adult (18-30yrs) subjects. The global CBF showed a monotonic decrease with age. After adjusting the global CBF differences, significant relative regional CBF increases with age were found in the cingulate, angular, superior temporal and frontal cortex, which may reflect the later maturation of these brain regions. These results support the feasibility and potential of ASL perfusion MRI for longitudinal neuroimaging studies to follow the functional development of the growing human brain.
Gary P. Zientara1, Janet S. Soul2, Kate McCann1, 3, Adre du Plessis2, Hiam Bassan2, Richard L. Roberston Jr. 2, Stephan E. Maier1, Ferenc A. Jolesz1, Lawrence P. Panych1, Steven A. Ringer1
1Harvard Medical School and Brigham and Women's Hospital, Boston, Massachusetts, USA; 2Harvard Medical School and Children's Hospital Boston, Boston, Massachusetts, USA; 3University of Maine at Orono, Orono, Maine, USA
Previously, we described a rapid EPI-based method for using slice inflow to investigate frequency band-specific pulsatile flow in the brain, and indicated its importance in revealing the physiology of the preterm infant. In a study of 30 preterm infants, we computed frequency band-specific pulsatile flow images, and then computed the distribution of flow-related signal. Our findings reveal that throughout the age range studied, the frequency distribution of pulsatile flow in the preterm infant brain is dominated by very low and low frequency components, the opposite of that described in published studies of flow in the adult brain.
2056.
PROPELLER FSE T2-Weighted Imaging in Pediatric Brain Imaging:
Can We Replace Standard FSE
T2-Weighted Imaging?
Alexandra Talia Vertinsky1, Erika Rubesova2, Michael Vladimir Krasnokutsky3, Sabine Bammer2, Jarrett Rosenberg2, Allan White2, Patrick David Barnes2, Roland Bammer2
1University of British Columbia, Vancouver, Canada; 2Stanford University, Stanford, California , USA; 3Madigan Army Medical Center, Tacoma, USA
PROPELLER is a relatively new method of reducing motion artifact, which produces similar motion reduction to SSFSE while providing improved image contrast. A comparative evaluation of T2w-FSE and T2w-PROPELLER on image quality and diagnostic yield was performed on a consecutive series of pediatric patients. Although no substantial benefit in diagnostic confidence was observed, T2w-PROPELLER FSE offers sufficient diagnostic information that it may be substituted for conventional FSE T2 in pediatric exams in order to reduce motion and improve image quality. However, caution is required in using PROPELLER to identify blood products and further studies are warranted to investigate the nature of this difference in sensitivity.
Hall D Tuesday 13:30-15:30
Efrat Sasson1, Glen M. Doniger2, Ofer Pasternak1, Yaniv Assaf1, 3
1Tel Aviv University, Tel Aviv, Israel; 2NeuroTrax Corporation, Newark, New Jersey, USA; 3Tel Aviv Sourasky Medical Center, Tel Aviv, Israel
The brain imaging correlates of age-related memory decline were studied in 51 healthy subjects, age 25-82 years. Diffusion imaging parameters were correlated with memory performance in multiple regions. Since these changes stem in part derived from age-related changes we conducted partial correlation between MRI parameters and memory performance controlling for age. W found that changes in brain integrity correlated with memory decline adjusted for age are most prominent in gray matter, in language-related regions (the STG, IFG) and hippocampal complex. These results suggest that diffusion MRI, especially the ADC, can be a useful regional quantitative marker for memory decline.
2091.
Orientation-Specific Degeneration of the Cerebral White
Matter in Aging Brain Investigated by
Geometric DTI
Chun-Jung Juan1, Yi-Jui Liu2, Cheng-Yu Chen1, Tzung-Tzuo Tsai2, Ming-Chung Chou3, Tzu-Cheng Chao3, Chun-Jen Hsueh1, Chung-Ping Lo1, Hsiao-Wen Chung3, Te-Cheng Lai2, Guo-Shu Huang1
1Tri-Service General Hospital, Taipei, Taiwan; 2Feng Chia University, Taichung, Taiwan; 3National Taiwan University, Taipei, Taiwan
Geometric DTI is capable of characterizing the diffusion anisotropy into linear, planar and spherical orientations. It is therefore an excellent tool to investigate the degeneration of the cerebral white matter. In this study, geometric DTI will show you the orientation-specific degeneration of the corpus callosum and periventricular white matter of the aging brain before the occurrence of volume loss. Specifically, the genu of corpus callosum and the anterior periventricular white matter demonstrate greater decrease of planar diffusion anisotropy, while the posterior periventricular white matter harbors greater loss of linear diffusion anisotropy in the elders.
2092. Robust Segmentation of White Matter Tracts in the Aging Brain
Mark E. Bastin1, Jonathan D. Clayden2, Jakub P. Piatkowski1, Amos J. Storkey1, Laura J. Brown1, Alasdair M J MacLullich1
1University of Edinburgh, Edinburgh, UK; 2University College London, London, UK
Understanding how aging affects brain structure is an important challenge for medical science. By allowing segmentation of tracts of interest from diffusion MRI data, tractography provides a promising tool for assessing white matter connectivity in old age. However, the output from tractography algorithms is strongly dependent on the subjective location of user-specified seed points, with the result that it can be difficult and time consuming to identify the same tract in different subjects. Here we evaluate the performance of a new method which reduces the sensitivity of tractography algorithms to seed point placement in the brains of normal aging subjects.
2093.
Combined Multi-Spectral Quantitative MRI and Volumetry of the
Brain with the Mixed-TSE Pulse
Sequence and Bisecting Dual-Clustering
Segmentation: A Technique for Studying Regional Ageing
Patterns in Large
Populations
Naoko Saito1, Osamu Sakai1, Hernan Jara1
1Boston Medical Center, Boston University School of Medicine, Boston, Massachusetts, USA
Purpose: To develop a combined quantitative MR imaging (Q-MRI) and volumetry technique for assessing simultaneously the segmental brain ageing patterns and volumetric changes in children, adults, and elder humans. Methods: The whole-head datasets of 75 subjects (0.5-87 years) generated with the mixed-TSE pulse sequence were segmented with a dual-clustering segmentation technique generating 6 sub-segments (bilateral frontal, bilateral posterior and bilateral cerebellar segments). Results: The volumetric and relaxometric age changes in the cerebellar segments were more gradual than in the posterior and frontal segments. Conclusion: A combined Q-MRI-volumetry technique has been developed and tested with 75 subjects over the human lifespan.
2094. Cortical Thickness and Mobility Status in Healthy Aging
Istvan Csapo1, Nicola Moscufo1, Leslie I. Wolfson2, Charles R. G. Guttmann1
1Brigham and Women's Hospital, Boston, Massachusetts, USA; 2University of Connecticut Health Center, Farmington, Connecticut, USA
We set out to investigate the link between cortical thickness and mobility status, white matter lesion load, and cognitive function in healthy elderly subjects. We found that regional cortical thickness correlates with mobility status. Thus, cortical thickness is a potentially important marker of mobility.
2095. Parallel Imaging in 3D MP-RAGE for Consistent Brain Volume Imaging
Ek Tsoon Tan1, Clifton R. Haider1, Roger C. Grimm1, Jeffrey L. Gunter1, Chadwick P. Ward1, Denise A. Reyes1, Matthew A. Bernstein1, Clifford R. Jack1, Stephen J. Riederer1
1Mayo Clinic, Rochester, Minnesota, USA
Serial bra