Traditional Posters : Pulse Sequences, Reconstruction & Analysis
Click on to view the abstract pdf and click on to view the pdf of the poster viewable in the poster hall.
New Methods for Generating Contrast

Tuesday May 10th
Exhibition Hall  13:30 - 15:30

2719.   Cube Cx2: Free 3D T2w dataset along with 3D T2FLAIR acquisition  
Donglai Huo1, and Xiaoli Zhao1
1GE Healthcare, Waukesha, WI, United States

3DFSE sequence (CUBE) is widely used in clinical now. In this abstract, a new 3DFSE acquisition method called Cube Cx2 (Cube Contrast times 2) is proposed to acquire a 3D Cube T2 weighted and a 3D Cube T2 FLAIR weighted data simultaneously, all within the same scan time as the original Cube T2 FLAIR sequence. Scan time are greatly saved (a 3D T2 weighted dataset is free), and the resulting two perfectly registered 3D datasets could also be potentially used for other research fields.

2720.   Variable Flip angle Single-slab 3D GRASE with Phase-independent Image Reconstruction 
Hahnsung Kim1, Suhyung Park2, Dong-Hyun Kim1, and Jaeseok Park3
1Electrical and Electronic Engineering, Yonsei University, Shinchon-Dong, Seoul, Korea, Republic of, 2Medical Science, Yonsei University, Seoul, Korea, Republic of, 3Radiology, Yonsei University, Seoul, Korea, Republic of

We develop flip-angle single-slab 3D GRASE imaging, incorporating multiple echo planar imaging (EPI) readouts into a framework of turbo/fast SE imaging, to enhance the imaging efficiency without the direct trade off with SNR. To avoid the phase discontinuity induced problems, phase-independent image reconstruction is performed, in which each echo k-space is regenerated by GRAPPA-like parallel imaging technique employing within-and between- group k-space correlations and then averaged to retain SNR efficiency.

2721.   High-Resolution 3D Volumetric Nerve-Sheath Weighted RARE Imaging (3D SHINKEI) 
Masami Yoneyama1, Masnobu Nakamura1, Tomoyuki Okuaki1, Takashi Tabuchi1, and Junko Ogura1
1Medical Satellite Yaesu Clinic, Tokyo, Japan

Magnetic resonance neurography is a useful technique with which to evaluate abnormal conditions of entire nerves and nerve bundles, and it has been used successfully in patients with tumors, trauma, and neuritis. In this study, we propose a new scheme of RARE (TSE) based fast high-resolution 3D volumetric peripheral nerve-sheath images using 3D nerve-SHeath signal increased with INKed rest-tissue RARE Imaging (3D SHINKEI). 3D SHINKEI technique combines fat-suppression prepulse, iMSDE@preparation for suppression from vessels, and 3D variable refocusing flip-angle RARE readout segments for contrast-efficient T2-weighted images. Furthermore, iMSDE prepulse was based on T2-prep@pulse; therefore, T2-contrast is affected by preparation duration.@We attempt to use that characterization for muscle@suppression. Consequently, 3D-SHINKEI technique was made possible to nerve-sheath depiction. 3D SHINKEI revealed detailed anatomy of lumbosacral plexus, brachial plexus, and cranial nerves. These results suggest that 3D SHINKEI can be used for fast high-resolution volumetric imaging of the peripheral nervous system.

2722.   Bipolar TSE and Bipolar 3D GRASE for Rapid Multi-slice (Multi-slab) High Field Magnetic Resonance Imaging Acquisition of Carotid Artery Wall 
Linqing Li1,2, and Peter Jezzard1,2
1FMRIB Centre, University of Oxford, Oxford, United Kingdom, 2Department of Clicical Neurosciences, University of Oxford, Oxford, United Kingdom

Flow-crushing bipolar gradient pairs are introduced into 2D TSE and 3D GRASE pulse sequences for black blood imaging multi-slice 2D or multi-slab 3D acquisition, termed as Bipolar 2D TSE and Bipolar 3D GRASE. Non-selective hard pulses used in conventional black blood modules are not required in these new imaging pulse sequences. Preliminary results indicate that the final average imaging speed has been increased dramatically up to 5 sec/slice or even faster (at least 2 or 3 times faster than conventional methods) with acceptable black blood imaging quality and comparable SNR.

2723.   Quiet T1- and T2-weighted brain imaging using SWIFT 
Ryan Chamberlain1, Steen Moeller1, Curt Corum1, Djaudat Idiyatullin1, and Michael Garwood1
1Center for Magnetic Resonance Research, University of Minnesota, Minneapolis, MN, United States

Sweep imaging with Fourier transform (SWIFT) is a novel 3D radial imaging technique that allows smooth and gradual gradient orientation changes during acquisition. Therefore, a SWIFT acquisition is nearly silent. A basic SWIFT sequence is proton density weighted due to low flip angles and very short echo times. It would be desirable to have a silent imaging technique that can create images with varying contrast. This work describes the use of SWIFT as a readout sequence with T1 and T2 preparation schemes, resulting in T1- and T2- weighted images acquired with minimal acoustic noise.

2724.   Magnetic resonance imaging of tendons, ligaments and menisci by subtraction of two steady state free precession signals 
Petros Martirosian1, Christina Schraml2, Nina Franziska Schwenzer2, Fabian Springer2, Fritz Schick1, and Michael Deimling3
1Section on Experimental Radiology, University of Tübingen, Tübingen, Germany, 2Department of Diagnostic and Interventional Radiology, University of Tübingen, Germany, 3Department of Magnetic Resonance, Siemens Healthcare, Erlangen, Germany

Structure and composition of certain musculoskeletal tissues result in short T2 values and nearly lacking signal using conventional imaging sequences. In this work we aimed at improving the visualization of tendons, ligaments, and menisci by subtraction of two steady state free precession signals which were simultaneously acquired in a constant readout gradient. By means of strong asymmetric echo acquisition and short non-selective excitation RF-pulse providing an echo time 0.8 ms it was possible to visualize tendons, ligaments and menisci as hyperintense structures in the subtraction images. The presented sequence is a promising approach which is worth being evaluated in clinical musculoskeletal MR imaging.

2725.   Quantitative MR estimates of blood oxygenation based on T2*: a numerical study of the impact of model assumptions. 
Thomas Christen1, Greg Zaharchuk1, Nicolas Pannetier2,3, Raphael Serduc2,3, Nicolas Joudiou2,3, Jean Claude Vial2,3, Chantal Remy2,3, and Emmanuel L Barbier2,3
1Department of radiology, Stanford University, Stanford, California, United States, 2U836, INSERM, Grenoble, France, 3Grenoble Institut des Neurosciences, Grenoble, France

Using numerical simulations, we analyzed the influence of assumptions of a mathematical model on estimates of T2*, blood volume fraction (BVf) and blood oxygen saturation (StO2). Our results suggest that the static dephasing regime (water diffusion neglected) is a good approximation as long as vessel radii are above 3 µm and StO2 is below 80% at 3T. MR estimates of StO2 obtained using the total BVf are in good agreement with the StO2 averaged over arterial and venous compartments. According to the results obtained using microscopy data as microvascular geometry input, using straight cylinders to model blood vessels seems appropriate.

2726.   Evaluation of a new quantitative BOLD approach to map local blood oxygen saturation in healthy rat 
Pierre Bouzat1,2, Thomas Christen1,3, Sébastien Thomas1,2, Nicolas Pannetier1,4, Chantal Rémy1,4, Jean-François Payen1,2, and Emmanuel L Barbier1,4
1U836, Inserm, Grenoble, France, 2CHU, Grenoble, France, 3Department of Radiology, Standford University, Standford, CA, United States, 4Grenoble Institut des Neurosciences, Université Joseph Fourier, Grenoble, France

To improve the accuracy and the spatial resolution of SO2MR maps obtained with quantitative BOLD approaches, an approach which combines a steady-state Blood Volume fraction measurement scheme, B0 and T2 mapping techniques, and a simpler mathematical model has recently been proposed. In this study, we evaluate the proposed acquisition scheme in healthy rats (n=17) while varying the inspired oxygen fraction and using a complete physiological control. We obtained a correlation between SO2MR and oxygen saturation measured in the sagital sinus of 0.84.

2727.   Is T2* Enough to Assess Oxygenation? A Quantitative Blood-Oxygen Level Dependent Analysis in Brain Tumors. 
Thomas Christen1, Benjamin Lemasson2,3, Nicolas Pannetier3,4, Regine Farion3,4, Chantal Remy3,4, Greg Zaharchuk1, and Emmanuel L Barbier3,4
1Department of radiology, Stanford University, Stanford, California, United States, 2Departments of Radiology, University of Michigan, Center for Molecular Imaging, Ann Arbor, Michigan, United States, 3Grenoble Institut des Neurosciences, Grenoble, France, 4U836, INSERM, Grenoble, France

We have analyzed how the addition of the transverse relaxation parameter (T2), the macroscopic field inhomogeneities (∆B0) and the blood volume fraction (BVf) to the estimates of magnetic resonance T2* influences the interpretation of BOLD oximetry results in a rat cerebral tumor model. We found no significant correlations between T2* and all other parameters in both tumor and healthy tissues. A lack of inclusion of any of these parameters may lead to incorrect assumptions about tumor oxygenation. Combining the parameters according to a quantitative BOLD approach leads to blood oxygenation (SO2) values independent of the initial T2* and consistent with previous measurements.

2728.   Measuring Brain Oxygenation in Humans using a Quantitative BOLD Approach 
thomas christen1, and Greg Zaharchuk1
1Department of radiology, Stanford University, Stanford, California, United States

We have recently developed an MR method to measure the level of blood oxygenation (SO2) in the brain. The technique is based on extraction of the oxygenation information from T2* measurements using a mathematical model and measurements of T2, B0 and the cerebral blood volume (CBV). Although the studies in healthy rat brain have shown good results, the steady state approach used for the determination of CBV is yet only applicable in rodents. The objective of this study was to translate the method for human use. Estimates of MR_SO2 were obtained on healthy volunteers by combining ASL and PWI measurements.

2729.   Rapid Measurement of Oxygen Extraction Fraction (OEF) Maps using a Combined Multiple Gradient and Spin Echo Bolus Contrast Sequence 
Thomas Christen1, Heiko Schmiedeskamp1, Matus Straka1, Roland Bammer1, and Greg Zaharchuk1
1Department of radiology, Stanford University, Stanford, California, United States

The objective of the study was to translate a quantitative BOLD approach originally designed to assess blood oxygenation in the rat brain into a clinical protocol. The technique is based on extraction of the oxygenation information from T2* measurements using a mathematical model and measurements of T2, B0 and the cerebral blood volume (CBV). We propose here to use an EPI multiple gradient and spin echo sequence to follow the bolus of a contrast agent. Baseline scans were used for T2*/T2 estimates whereas PWI was used to derive CBV maps. 4 healthy subjects and one stroke patient were scanned.

2730.   Evaluation of a new qBOLD approach to map local blood oxygen saturation in human brain 
Julien Y BOUVIER1,2, Irène TROPRES3,4, Marjorie VILLIEN1,5, Sylvie GRAND1,6, Assia JAILLARD4,6, Omer EKER3,6, Olivier DETANTE1,6, David CHECHIN2, Jean-François LE BAS3,6, Alexandre KRAINIK1,6, and Emmanuel L BARBIER1,5
1Grenoble Institut des Neurosciences, Université Joseph Fourier, Grenoble, France, 2Philips Healthcare, Suresnes, France, 3Université Joseph Fourier, Grenoble, France, 4IFR1, Grenoble, France,5U836, INSERM, Grenoble, France, 6CHU, Grenoble, France

A simple MR approach for local oxygen saturation (lSO2) assessment in human brain at 3T is evaluated. It combines separate estimates of T2, T2*, Blood Volume fraction (BVf) and B0 inhomogeneities. lSO2, BVf, T2 and T2* maps of four patients are presented. Mean lSO2 values in gray matter, white matter or both were 57±5%, 34±3%, and 45±4% respectively. This study shows that cerebral lSO2 may be measured with good spatial resolution in a short MR exam using three MR sequences. Further studies are required to validate this promising approach in human and improve the estimates obtained in white matter.

2731.   Acoustic radiation contrast to visualize viscoelastic properties in human breast 
Deniz Ulucay1, Judith Wild1, Jessica Mende2, Michael Dönnebrink3, Jürgen Finsterbusch4, Carsten Urbach1, and Karl Maier1
1HISKP, University of Bonn, Bonn, NRW, Germany, 2Lavandoo Mobile Solutions GmbH, Bonn, NRW, Germany, 3Medizin Center Bonn, Bonn, NRW, Germany, 4University Medical Center Hamburg-Eppendorf, Hamburg, Germany

Acoustic radiation contrast in magnetic resonance (ARC-MR) phase images is an innovative and recently developed method to visualize lesions and microcalcifications non-invasively based on differences in viscoelastic properties. Acoustic radiation force was applied using a custom made MR compatible piezoelectric transducer with a resonance frequency of 2.5 MHz. The thus produced displacement in breast tissue was made visible with a displacement sensitive spin-echo sequence. Measurements on three healthy volunteers show the feasibility of ARC-MR in human breast. It is expected that this method benefits breast cancer diagnostics by providing viscoelastic properties.

2732.   Towards direct Neuronal Current Imaging by Resonant Rabi Oscillation Mechanisms 
Alexey Tonyushkin1, and Andrew M Kiruluta1,2
1Physics, Harvard University, Cambridge, Massachussetts, United States, 2Radiology, MGH, Boston, Massachussetts, United States

Current approaches aimed at understanding brain function can be broadly divided into those that rely on hemodynamic responses as indicators of neural activity (fMRI, Optical and PET) and methods that measure neural activity directly (MEG and EEG). These approaches all suffer from poor temporal resolution (fMRI), poor spatial localization (MEG and EEG), or indirectly measuring neuron activity (fMRI, Optical and PET). It has been suggested that the proton spin population will be altered by neural activity fields resulting in a measurable effect on the MR signal that can be imaged by MRI methods. Unfortunately, this effect has been determined to be too small to be detectable. We present the physical basis and experimental evidence for an alternative approach based on a resonant interaction between the magnetic fields such as those arising from neuron activity, with a spin population that is undergoing Rabi oscillations at a frequency commensurate with the neuron currents. It is well established that neural firing during an activation has a spectrum associated with it.

2733.   Highly efficient localized distant dipolar field and its application in MRI 
Congbo Cai1, Zhong Chen1, Shuhui Cai1, and Jianhui Zhong2
1Departments of Physics and Communication Engineering, Xiamen University, Xiamen, Fujian, China, People's Republic of, 2Departments of Radiology and Biomedical Engineering, University of Rochester, Rochester, United States

Intermolecular multiple quantum coherences (iMQCs) possess some appealing unique properties for MRI. However, their intrinsic low signal to noise ratio prevents their widespread application. In this abstract, a new method was designed to produce a highly efficient localized distant dipolar field, thus substantially enhancing the iMQC signal intensity. Its application in MRI was tested. Experimental and simulation results indicate that the iMQC signal intensity is enhanced by more than two folds in comparison with the conventional CRAZED method.

2734.   Reference free Localization and Quantification of Contrast Agents using Relaxivity Dispersion at 1.5T 
Uvo Christoph Hoelscher1, Steffen Lother1, Florian Fidler1, and Peter Jakob1,2
1Research Center Magnetic Resonance Bavaria (MRB), Wuerzburg, Bavaria, Germany, 2Department for Experimental Physics 5 (Biophysics), University of Wuerzburg, Wuerzburg, Germany

Many medical problems can benefit from unambiguous localization and quantification of contrast agents, but conventional detection / quantification methods always require reference scans to separate contrast agent and tissue. In this work we present a new method to acquire both information without the need for a reference scan. The technique exploits relaxivity dispersion of the contrast agent making separation between contrast agent and tissue unambiguous. Resulting images show pure contrast agent signal and can be used to determine concentrations.

2735.   Exchange-Relayed Nuclear Overhauser Effect MRI 
Craig Kenneth Jones1,2, Alan J Huang1,3, and Peter C M van Zijl1,2
1FM Kirby Center, Kennedy Krieger Institute, Baltimore, MD, United States, 2Department of Radiology and Radiological Sciences, Johns Hopkins Medical Institutes, Baltimore, MD, United States,3Department of Biomedical Engineering, Johns Hopkins University, Baltimore, MD, United States

Recently, chemical exchange saturation transfer (CEST) MRI has been used to detect the signals of mobile proteins and peptides in vivo through the exchange between amide protons and water protons. It is well known that mid to large size mobile macromolecules should experience cross-relaxation effects called Nuclear Overhauser Enhancements (NOEs), the build-up of which should be slower than exchange effects. In this work, exchange-relayed NOEs of mobile proteins are shown to be visible in CEST images in vivo. This is possible because a small B1 irradiation field was used to avoid competing effects from exchange and semi-solid magnetization transfer effects.

2736.   Observation of intravascular contrast enhancement due to anesthesia in T2*-weighted imaging at 17.2 T 
Luisa Ciobanu1, Olivier Reynaud1, Béchir Jarraya2, and Denis Le Bihan1
1NeuroSpin, CEA, Gif-sur-Yvette, France, 2NeuroSpin, INSERM-A VENIR unit, Gif-sur-Yvette, France

T2*-weighted imaging at ultra high magnetic field strengths benefits from a dramatic increase in contrast to noise ratio. We show that the vessels/tissue contrast in T2*-weighted images at 17.2 T is greatly influenced by the anesthetic agent used. Stemming from magnetic susceptibility differences, this phenomenon is visible to a much smaller extent at lower field strengths. We present high contrast T2* -weighted images of rodent brain microvasculature acquired in vivo, at 17.2 T under certain anesthesia conditions in the absence of contrast agents.

2737.   The use of Iteratively Reweighted Least Square (IRLS) in the calculation of tissue susceptibility 
Tian Liu1,2, Cynthia Wisnieff1,2, Craig Horenstein3, Krishna Surapaneni3, and Yi Wang1,2
1Biomedical Engineering, Cornell University, Ithaca, NY, United States, 2Radiology, Weill Cornell Medical College, New York, NY, United States, 3Radiology, Columbia University, New York, NY, United States

Calculation of susceptibility from measured field map is an ill-posed inverse problem. In addition, the noise on the field map may not be Gaussian in signal void regions or due to improper phase unwrapping, complicating the inversion. In this abstract, we propose to use an Iteratively Reweighted Least Square (IRLS) algorithm to automatically identify the outlier pixels with non-Gaussian field noise, and attenuate the weighting for these pixels. Preliminary results showed that IRLS is able to suppress streaking artifacts in the presence of clusters of outlier pixels.

Traditional Posters : Pulse Sequences, Reconstruction & Analysis
Click on to view the abstract pdf and click on to view the pdf of the poster viewable in the poster hall.
Relaxometry I

Wednesday May 11th
Exhibition Hall  13:30 - 15:30

2738.   Improved Single-shot MR Relaxometry using Principal Component Analysis 
Philipp Ehses1, Xavier Helluy1, Michael Völker1, Vikas Gulani2, Nicole Seiberlich2, Mark A Griswold2, Peter M Jakob1,3, and Felix A Breuer1
1Research Center for Magnetic Resonance Bavaria (MRB), Würzburg, Germany, 2Dept. of Radiology, University Hospitals of Cleveland and Case Western Reserve University, Cleveland, United States, 3Dept. of Experimental Physics 5, Universität Würzburg, Würzburg, Germany

Quantitative MRI has not yet found widespread clinical adoption, mainly due to long scan times required to map the various parameters of interest. PCA has previously been used to reconstruct data from an undersampled time series, and it has been demonstrated that relaxometry dynamics can be well described by a small number of principal components. In this work, we present a novel PCA based reconstruction method, which allows the quantification of T1, T2 and spin-density from a single IR TrueFISP experiment in under 5s per slice. Undersampling artifacts were completely removed and the results were comparable to a fully-encoded reference.

2739.   Improving T2* mapping at 7 T 
Weiqiang Dou1, Ralf Deichmann2, Oliver Speck1, and Kai Zhong1
1Biomedical Magnetic Resonance, Otto-von-Guericke University, Magdeburg, Saxon-Anhalt, Germany, 2Brain Imaging Center, Johann Wolfgang Goethe-University Frankfurt/M., Frankfurt/Main, Hesse, Germany

Strong B0 field inhomogeities at 7T can decrease significantly brain T2* values that are related to brain tissue iron content. An improved correction method considering both linear and quadratic B0 variations was described and implemented in this study. The results showed significant improvement on T2* values compared to that of the linear correction method at intermediate susceptibility gradients (50 – 150 lower case Greek muT/m). The new method therefore allow for more reliable T2* mapping for 7T in vivo studies.

2740.   Accelerating Multi-Component Relaxometry in Steady State with an Application of Constrained Reconstruction in Parametric Dimension 
Julia V Velikina1, Samuel A Hurley1, Andrew L Alexander1, and Alexey A Samsonov1,2
1Medical Physics, University of Wisconsin - Madison, Madison, WI, United States, 2Radiology, University of Wisconsin - Madison, Madison, WI, United States

We propose a novel way to accelerate multi-component relaxometry by a factor of 4 by applying variable density undersampling in the flip angle dimension of the acquired SPGR and bSSFP data and then reconstructing the obtained incomplete data sets using constrained reconstruction in the parametric (flip angle) dimension. Finally, parameter maps, such as myelin water fraction, are derived from the reconstructed image series. We compare our results with the ones obtained using parallel imaging alone and fully sampled data.

2741.   Nonlinear Inverse Reconstruction for T2 Mapping from Highly Undersampled Cartesian Spin-Echo MRI 
Tilman Johannes Sumpf1, Martin Uecker1, Susann Boretius1, and Jens Frahm1
1Biomedizinische NMR Forschungs GmbH, Goettingen, Germany

Quantitative evaluations of the T2 relaxation time are of high importance for diagnostic MRI. Standard T2 mapping procedures rely on the time-demanding acquisition of fully sampled k-space datasets at multiple echo times. Recently, a nonlinear inverse reconstruction method has been proposed which allows for reconstructions of spin-density and T2 maps from highly undersampled radial data. Here, we extended this concept for Cartesian acquisition schemes and even for settings without parallel imaging. An automatic scaling procedure as well as a dedicated undersampling pattern minimize residual artifacts. The approach is validated for T2 mapping of a numerical phantom and the human brain.

2742.   Average Correlation Orthogonal Matching Pursuit for Improved Relaxation Parameter Estimation 
Nicole Seiberlich1, Dan Ma2, Philipp Ehses3, Vikas Gulani1, and Mark Griswold1,2
1Radiology, University Hospitals of Cleveland, Cleveland, OH, United States, 2Biomedical Engineering, Case Western Reserve University, Cleveland, OH, United States, 3Research Center for Magnetic Resonance Bavaria (MRB), Wuerzburg, Germany

Orthogonal Matching Pursuit (OMP) has emerged as a powerful tool for the quantification of MRI relaxation parameters, where a dictionary of simulated signal evolution curves (atoms) is compared with undersampled images in order to determine relaxation parameter maps. When working with multiple parameters (quantifying T1, T2, and M0 simultaneously), atoms can be quite similar, making OMP quantification inaccurate. Using an approach that considers the average correlation instead of simply the highest correlation, we show that the relaxation parameters can be more accurately determined. Average correlation OMP is demonstrated using IR-TrueFISP to yield accurate T1 and T2 maps in abdominal imaging.

Traditional Posters : Pulse Sequences, Reconstruction & Analysis
Click on to view the abstract pdf and click on to view the pdf of the poster viewable in the poster hall.
Relaxometry II

Thursday May 12th
Exhibition Hall  13:30 - 15:30

2743.   Saturation Recovery Modified Look Locker (S-MOLLI) for Cardiac T1 Mapping 
Christian Stehning1, Daniel Messroghli2, Michael Frick3, Bernhard Schnackenburg3, and Jochen Keupp1
1Philips Research Laboratories, Hamburg, Germany, 2Cardiac MRI Unit, Franz-Volhard-Klinik, Charité University Medicine, Berlin, Germany, 3Department of Internal Medicine/Cardiology, German Heart Institute, Berlin, Germany

A saturation recovery variant of modified Look Locker (MOLLI) for cardiac T1 mapping is presented. It offers reduced scan times and invariant T1 values over a wide range of RR intervals.

2744.   A universal sampling scheme for the Method of Slopes (MoS) allows for rapid simultaneous B1 and T1 mapping in 2D 
Sofia Chavez1, and Greg Stanisz1,2
1Imaging Research, Sunnybrook Research Institute, Toronto, ON, Canada, 2Medical Biophysics, University of Toronto, Toronto, ON, Canada

The Method of Slopes (MoS) has been proposed for simultaneous 3D B1 and T1 mapping. It uses 3D SPGR signal at different nominal flip angles (FAs). An extrapolation to signal null is used to obtain B1. This extrapolation is valid if the sampled signal varies linearly with FA but the region of linearity varies spatially with B1. In this work, a method is proposed to overcome the inaccuracies in the extrapolation when the signal is not sampled in the linear region resulting in a universal sampling scheme which allows for a 2D signal calibration, extending the MoS to 2D applications.

2745.   Impact of three B1 mapping techniques on variable flip angle T1 measurements 
Christine Lucas Tardif1, Nikola Stikov1, Ives R Levesque2, and G Bruce Pike1
1McConnell Brain Imaging Centre, Montreal Neurological Institute, Montreal, Quebec, Canada, 2Department of Electrical Engineering, Stanford University, Stanford, California, United States

Variable flip angle (VFA) T1 mapping has become a popular tool to estimate T1 times in vivo due to its time-efficient high-resolution 3D coverage. For accurate T1 estimates at 3 Tesla, the acquisition of a B1 map is essential to correct the nominal flip angles. This work evaluates the impact of three published B1 mapping techniques (the 2D double angle method (DAM), 3D actual flip angle imaging (AFI) and 3D Bloch-Siegert shift (BS)) to correct VFA T1 measurements in phantoms. Optimally spoiled VFA T1 mapping with DAM B1 correction yields the most accurate results. The errors in AFI and BS B1 maps result in broadening and/or shifting of the T1 histograms.

2746.   Accelerated T1 and T2 Relaxometry in the Human Brain Using UNFOLD 
Ana-Maria Oros-Peusquens1, and Nadim Jon Shah1,2
1Institute of Neuroscience and Medicine (INM-4), Research Centre Juelich, Juelich, NA, Germany, 22Department of Neurology, Faculty of Medicine, JARA, RWTH Aachen University, Aachen, Germany

We present an application of the UNFOLD method to T1 and T2 relaxometry which takes advantage of the high temporal resolution of two multi-point mapping methods; the inherent and extensive temporal information is used for acquisition time reduction. A multi-time-point, multi-slice Look-Locker sequences, TAPIR, was used for T1 mapping and a conventional CPMG-type multi-slice multi-echo sequence for T2 mapping. Comparing full k-t space and undersampled k-t space, the mean value over the brain is below 3%; the undersampled acquisition is, however, twice as fast.

2747.   Mapping of Oxygen By Imaging Lipids relaxation Enhancement (MOBILE): Application to Changes in Liver Oxygenation 
Benedicte F Jordan1, Julie Magat1, Elif Ozel1, Valerie Marchand1, Patrice Cani2, Nathalie Delzenne2, and Bernard Gallez1
1Louvain Drug Research Institute, Biomedical Magnetic Resonance Research Group, University of Louvain, Brussels, Belgium, 2Louvain Drug Research Institute, Metabolism and Nutrition Research Group, University of Louvain, Brussels, Belgium

In the present study, we propose to exploit the higher solubility property of oxygen in lipids than in water to monitor the changes in R1 of the lipid peak and translate it into pO2 values. For this purpose, we developed a sequence that is able to map variations in oxygenation based the relaxation properties of the tissue lipids. We measured in vitro the relaxation properties of water and lipid components in pure aqueous or oil phases, and tissue homogenates equilibrated in different oxygen environments, and monitored the evolution of the R1 of lipids in vivo in the liver of mice before and during a carbogen breathing challenge. The measurement of R1 in lipids offers an increased sensitivity when monitoring the changes in tissue oxygenation compared to previously described techniques that measure the variations of R1 in the water component.

2748.   Optimization Strategies for Relaxation based Myelin Water Imaging: 2. Postprocessing and Signal Correction Techniques 
Burkhard Mädler1, and Volker A. Coenen1
1Dep. of Neurosurgery, Div. of Stereotaxy and MR-based OR-Techniques, University Bonn, Bonn, Germany

We show the validity and usefulness of three retrospective correction techniques for multi-spin echo myelin water imaging (MWI) that do not generally require additional data acquisition: a dynamic T2- integration window method for the reliable and robust assessment of the myelin peak, an algorithm to successfully correct for stimulated echo artifacts in the T2-echo decay curve and a complex real-data rephasing approach to eliminate baseline offsets due to non-Gaussian noise in the decay curve. They lead to substantially better estimations of quantitative myelin fraction values even under the presence of large B1-inhomogeneities and variations in physiological noise of the data.

2749.   Improved T2-Quantification with Slice Selective MSE-Sequences 
Andreas Petrovic1,2, Eva Scheurer2, Kathrin Yen2, and Rudolf Stollberger1
1Institute of Medical Engineering, University of Technology Graz, Graz, Austria, 2Ludwig Boltzmann Institute - Clincal Forensic Imaging, Graz, Austria

Fast and accurate measurement of the transverse relaxation time T2 has been the goal of numerous papers. However, in the presence of B1+ inhomogeneities and non-ideal slice profiles T2quantification is impaired by T1 mixing effects. This leads to an overestimation of T2 when exponential curve fitting is used. In this work the generating functions approach is used to calculate a realistic decay which can be fitted to the multi-echo data. The algorithm was validated on phantom and in-vivo data. T1 mixing could be strongly reduced and accurate T2 values can be obtained.

2750.   Monte Carlo analysis of T1-mixing errors for MSE T2 mapping 
Andreas Petrovic1,2, Eva Scheurer2, Kathrin Yen2, and Rudolf Stollberger1
1Institute of Medical Engineering, University of Technology Graz, Graz, Styria, Austria, 2Ludwig Boltzmann Institute - Clincal Forensic Imaging, Graz, Austria

Quantitative MRI aims at providing a comparable measure of tissue condition by measuring physical properties. However, in reality the purpose of qMRI is limited by technical restrictions. T2mapping with multi spin echo sequences is usually impaired by T1-mixing effects. The actual exponential T2 decay is altered by contributions from stimulated echoes. In this study an error analysis has been carried out considering the mono-exponential model and an elaborate generating functions model that accounts for the systematic error. Simulations indicate that using the mono-exponential model yields largely overestimated T2 values compared to the generating functions approach.

2751.   Robust multicomponent T2 imaging in the brain at 3 T using least squares fitting in the presence of RF inhomogenities 
Sha Zhao1,2, David L Buckley3, and Geoff JM Parker1,2
1Imaging Science, The University of Manchester, Manchester, United Kingdom, 2Biomedical Imaging Institute, The University of Manchester, Manchester, United Kingdom, 3Division of Medical Physics, University of Leeds, Leeds, United Kingdom

We propose a method for multicomponent T2 imaging of the brain at 3 T using least squares fitting for turbo-spin-echo imaging data. Commonly multiple echo data of a CPMG sequence are acquired then NNLS transformed to generate T2 spectrum images. At 3 T RF inhomogeneity affects magnetisation refocusing pulses, leading to contamination by stimulated echo signals, which degrades the generated T2 spectrum images. We present a least squares fitting method that takes into consideration the signal contribution from stimulated echoes and is capable of generating T2 component images that are less affected by RF field inhomogeneity.

2752.   A Simple Method for Increasing the Number of Echoes and Decreasing Echo Spacing in T2 Spectrum Analysis 
Marshall S Sussman1
1Medical Imaging, University Health Network, Toronto, Ontario, Canada

A key requirement of measuring T2 spectra is high quality data. Specifically; high SNR, a large number of echoes, and short echo spacing. Most in vivo implementations achieve high data quality exclusively through the use of multiple averages to increase SNR. In this study, we present a very simple technique that also uses multiple acquisitions to improve data quality. However, the novel feature of this technique is that multiple acquisitions are used to allow for more flexible echo spacing and a larger number of echoes. This additional flexibility may bring the promise of T2 spectrum analysis closer to clinical reality.

2753.   Measuring and Imaging T2 without Echoes? 
Guan Wang1,2, AbdElMonem M. El-Sharkawy1,2, William A. Edelstein1, Michael Schär1,3, and Paul A. Bottomley1,2
1Radiology and Radiological Science, Johns Hopkins University School of Medicine, Baltimore, Maryland, United States, 2Electrical and Computer Engineering, Johns Hopkins University, Baltimore, Maryland, United States, 3Philips Healthcare, Ohio, Cleveland, United States

T2 is conventionally measured by the spin-echo (SE) method. Here we present a new method of measuring T2 without echoes, utilizing the fact that long-duration adiabatic excitation pulses are prone to T2 decay during excitation. T2 is measured from the ratio of MR signals acquired with and without long-duration adiabatic pulses. T2-weighted MRI is performed by incorporating a long-duration adiabatic pre-pulse in the MRI sequence. Use of 0° adiabatic pre-pulses ensures that the other contrast properties of the MRI sequence are unaffected. The method is validated on phantoms by comparison with the SE method, and reduced to practice in MRI.

2754.   Multi-parameter mapping of the human cervical cord at 3.0T in less than 20 minutes 
Rebecca Sara Samson1, Olga Ciccarelli2, Carolina Kachramanoglou2, Antoine Lutti3, David J L Thomas2, Nikolaus Weiskopf3, and Claudia A M Wheeler-Kingshott1
1Department of Neuroinflammation, UCL Institute of Neurology, London, England, United Kingdom, 2Department of Brain Repair and Rehabilitation, UCL Institute of Neurology, London, England, United Kingdom, 3Functional Imaging Laboratory, UCL Institute of Neurology, London, England, United Kingdom

Quantitative MRI techniques have been shown to be sensitive to changes in diseased spinal cord via measurements of relaxation times, the Magnetisation Transfer Ratio (MTR), and more recently quantitative Magnetisation Transfer (MT) parameters. We have developed an imaging protocol to enable mapping of proton density (PD), T1, MTR, MT and R2* in under 20 minutes. Here we present data acquired in the spinal cords of 12 healthy volunteers using this rapid multi-parameter mapping protocol. This technique may provide an insight into pathological processes occurring in the cervical cord affected by neurological disorders such as Multiple Sclerosis (MS).

2755.   A new 3D method for water and relaxation time mapping: comparison to the 2D “gold standard” 
Ana Maria Oros-Peusquens1, Fabian Keil1, Vincent Gras1, Zaheer Abbas1, Daniel Brenner1, Miriam Rabea Kubach1, Klaus Hans Mannfred Möllenhoff1, and Nadim Jon Shah1,2
1Institute of Neuroscience and Medicine - 4, Forschungszentrum Juelich, Juelich, Germany, 2Department of Neurology, Faculty of Medicine, JARA, RWTH Aachen University, Aachen, Germany

A new 3D method for water content and relaxation time mapping in which the M0 and T1 are determined from a two-point fit to measured data sets is presented. The method is based on 3D multiple-echo gradient echo acquisitions and thus also allows for accurate determination of T2*. The new method is applied at 3T and compared to a “gold standard” 2D method with very long TR. Results from phantoms and 5 healthy subjects are presented. The accuracy of the 3D method is better than 5%.

2756.   Evaluation of Principal Component Model-based algorithm for T2 estimation of small objects 
Chuan Huang1, Christian G Graff2, Ali Bilgin3,4, and Maria I Altbach5
1Mathematics, University of Arizona, Tucson, Arizona, United States, 2Division of Imaging and Applied Mathematics, Food and Drug Administration, 3Biomedical Engineering, University of Arizona,4Electrical and Computer Engineering, University of Arizona, 5Radiology, University of Arizona

Radial Fast Spin Echo (FSE) methods have been proposed for fast T2 mapping using highly undersampled data. One approach to process the data is based on echo sharing where mixed TE data sets are used to reconstruct T2 maps. Due to the mixing of TE data, however, T2 estimation for small structures could be problematic. Recently a Principal Component Model-based algorithm was developed to reconstruct TE images from highly undersampled radial FSE data. The method utilizes principal component coefficient maps to reconstruct TE images without mixing TE information. In this work, we evaluate the model-based algorithm based on the accuracy of T2 estimation and compare it to the echo sharing approach.

2757.   Spatially Resolved Two-Dimensional T1-T2 Relaxometry in the Human Brain Using Inversion-Recovery Spin-Echo Measurements and NNLS 
Valentin Gereon Kemper1, Ana-Maria Oros-Peusquens1, and Nadim Jon Shah1,2
1Institute of Neuroscience and Medicine, Research Centre Juelich, Juelich, 52425, Germany, 2Department of Neurology, Faculty of Medicine, JARA, RWTH Aachen University, Aachen, 52074, Germany

Spatially resolved and correlated T1-T2 distribution spectra were obtained pixelwise in a human post mortem brain slice using multiple inversion recovery spin echo scans and 2D nonnegative least squares for data analysis. This way multiple T1 compartments can be associated with their respective T2 values. Resolution was 0.75x0.75x2mm3. The obtained spectra show characteristic distributions for different types of tissue yielding sub-pixel scaled information. The width of the white matter spectra is much wider in both, T1 and T2, supporting the assumption of multiple water compartments shown. Short and long component maps were obtained using both T1 and T2 distributions and their combination.

2758.   Fast proton density mapping using bias field correction 
Steffen Volz1, Ulrike Nöth1, and Ralf Deichmann1
1Brain Imaging Center (BIC), Goethe University Frankfurt, Frankfurt, Germany

A method for fast mapping of the proton density (PD) is presented, based on advanced analysis of data obtained with the variable flip angle technique. In particular, corrections for the sensitivity profile of the receiver coil are performed by creating a synthetic T1 weighted anatomical image where all other sources of signal non-uniformities have been removed. Thus, correction factors can be directly deduced from the bias field following from unified segmentation of the synthetic anatomy. Numerical PD values obtained from six subjects are in very good agreement with literature values.

2759.   A Novel Method for Characterizing T2 Spectra 
Marshall S Sussman1, and Walter Kucharczyk1
1Medical Imaging, University Health Network, Toronto, Ontario, Canada

Measurement of a tissue’s T2 spectrum provides information beyond that of methods used in current clinical practice. However, T2 spectrum analysis is rarely used clinically. This is because the current method used to measure T2 spectra, multi-exponential fitting, requires very long scan times. An alternate method for assessing T2 spectra is linear combination filtering (LCF). LCF can be performed in clinically reasonable scan times. However, it only assesses a single, extended region of the T2 spectrum. In this project, we develop a novel LCF technique that provides an estimate of the full T2 spectrum in a clinically reasonable scan time.

2760.   Comparison of Two MRI-UTE Sequences for the Quantification (T1) of the Human Achilles Tendon 
Peter Wright1, Richard Hodgson2, Vladimir Jellus3, Lars Lauer3, and Matthew Robson4
1LMBRU, Leeds Teaching Hospitals NHS Trust, Leeds, Yorkshire, United Kingdom, 2LMBRU, University of Leeds, United Kingdom, 3Siemens AG, Erlangen, Germany, 4University of Oxford, United Kingdom

The Achilles tendon is commonly involved in degenerative and inflammatory tendinopathies. The aim was to compare T1 measurements in healthy human Achilles tendon using a saturation recovery (SR-) UTE and variable flip angle (VFA-) UTE sequence, which were initially corroborated in phantoms with an assumed ‘gold standard’ inversion recovery spin echo (IR-SE) sequence. Phantom results showed good comparison between IR-SE and SR-UTE sequences (r2 = 0.9985 [p<0.02]) and IR-SE and VFA-UTE sequences (r2 = 0.9993 [p<0.01]). Healthy volunteer T1 measurements in the Achilles tendon were 725 ± 42 ms and 698 ± 54 ms for SR- and VFA-UTE sequences respectively.

2761.   Transverse Relaxometry with non-180° Refocusing Pulses 
Julien Sénégas1, Nicolas Neu2, and Jochen Keupp1
1Philips Research Laboratories, Hamburg, Germany, 2Ecole des Mines de Paris, France

Recently, a model based on the extended phase graph (EPG) approach has been proposed to fit the signal curve of multi-echo spin-echo acquisitions pulses with the goal to improve the robustness of T2 estimation with respect to stimulated echoes. While the EPG approach has been proposed primary to correct for B1 inhomogeneities and imperfect refocusing pulses, we investigate here whether this approach can be applied with sufficient accuracy and precision in the case of pulses with refocusing angles below 180˚ with the benefit of reduced power deposition and shorter echo spacing.

2762.   A Four Parameter Fitting Method to Quantify Fully the Sources of Phase Contrast in Gradient Echo MRI 
Sam Wharton1, and Richard Bowtell1
1Sir Peter Mansfield Magnetic Resonance Centre, University of Nottingham, Nottingham, United Kingdom

Until recently, the excellent contrast seen in gradient echo based phase images was thought to be entirely due to differences in the isotropic magnetic susceptibility of tissues. Recently, however, three additional contrast mechanisms have been proposed. These are: (i) exchange processes; (ii) field perturbations due to oriented NMR-invisible microstructures; (iii) anisotropy of the magnetic susceptibility. Here, we present a method for fitting and separating the contributions of each of the four proposed contrast mechanisms using MRI phase data acquired with the sample at multiple orientations to the main magnetic field, B0, in combination with diffusion tensor imaging data.

2763.   Region Based Joint Bi-exponential T2 Fitting for Small Lesions 
Chuan Huang1, Christian G Graff2, Eric W Clarkson3,4, Ali Bilgin5,6, and Maria I Altbach3
1Mathematics, University of Arizona, Tucson, Arizona, United States, 2Division of Imaging and Applied Mathematics, Food and Drug Administration, 3Radiology, University of Arizona, 4Optical Sciences, University of Arizona, 5Biomedical Engineering, University of Arizona, 6Electrical and Computer Engineering, University of Arizona

T2 estimation has proven to be a valuable quantitative tool for assessing a variety of pathologies. It plays a particularly important role in lesion classification. Partial volume is generally a factor in lesion T2 estimation, particularly for lesions with diameters smaller than 15 mm. In order to obtain accurate T2 estimates for lesions with partial volume, bi-exponential fitting is required. However, bi-exponential model fitting suffers from large uncertainty of the fitting parameters when noise is present. In this work, we propose a novel ROI-based joint bi-exponential fitting algorithm to estimate T2 of lesions affected by partial volume. This approach takes advantage of the lesion fraction variation among voxels within an ROI (which is naturally present in small lesions) to improve T2 estimation. The performance of the ROI-based joint bi-exponential fitting algorithm is evaluated in simulations and real MRI data.

2764.   Accurate T1 and T2 Quantification in Look-Locker 2D SSFP Imaging with Flip Angle Profile Correction 
Mitchell Anthony Cooper1,2, Thanh D Nguyen2, Pascal Spincemaille2, Martin R Prince2, Jonathan W Weinsaft3, and Yi Wang1,2
1Biomedical Engineering, Cornell University, Ithaca, New York, United States, 2Radiology, Weill Cornell Medical College, New York, New York, United States, 3Cardiology, Weill Cornell Medical College, New York, New York, United States

In this study we investigated the accuracy of a new method to fit Look-Locker 2D SSFP data with flip angle profile correction. Utilizing the bloch equations and integrating over the flip angle profile, we were able to determine T1 with substantially lower error when compared to previous methods for SSFP T1 mapping sequences. We were able to accurately measure T1 and T2 in the thigh on a healthy volunteer cohort.

2765.   In vivo T2 Measurements of the Right Ventricle Inferior Wall: Comparison with the Left Ventricle 
Brice Fernandez1,2, Maelene Lohezic1,2, Lucien Hammen2,3, Marine Beaumont4,5, Damien Mandry2,4, Pierre-André Vuissoz2,3, and Jacques Felblinger2,3
1Global Applied Science Laboratory, GE Healthcare, Nancy, France, 2IADI Lab, Nancy-Université, Nancy, France, 3U947, INSERM, Nancy, France, 4CHU de Nancy, Nancy, France, 5CIT801, INSERM, Nancy, France

Transverse relaxation time T2 of the right ventricle (RV) wall is not a well known quantity and its value has only been reported by few studies, most likely due to the thin right ventricle wall avoiding accurate measurements. In this study, T2 of the RV inferior wall was measured in end-systolic rest, when RV wall is thickened, using a recently described adaptive method and a double inversion recovery fast spin echo sequence on a 1.5T scanner. T2 values of the RV in healthy volunteers were significantly longer than those of the LV which is in concordance with other studies.

Traditional Posters : Pulse Sequences, Reconstruction & Analysis
Click on to view the abstract pdf and click on to view the pdf of the poster viewable in the poster hall.
Chemical Exchange Saturation Transfer

Monday May 9th
Exhibition Hall  14:00 - 16:00

2766.   Feasibility of CEST imaging on the guinea pig stifle at 9.4 T 
Matthew Fenty1, Victor Babu Kassey1, Feliks Kogan1, and Ravinder Reddy1
1CMROI, Radiology, University of Pennsylvania, Philadelphia, PA, United States

Osteoarthritis (OA) is a common and painful condition with a multi-factorial etiology of the musculoskeletal system affecting more than 50% of the U.S. population over 65. Degeneration of the articular cartilage tissue is a slow process and typically takes decades to develop. Dunkin-Hartley guinea pigs have been shown to develop OA with the earliest stage of detection manifesting as early as three to four months of age. Chemical Exchange Saturation Transfer (CEST) has become a popular method for measurement of metabolites with exchangeable protons. GAG contains amino groups and hydroxyl groups that exchange protons with bulk water that can be that exploited for CEST. Here we will show preliminary results of implementation and optimization of CEST imaging using GAG-dependent contrast on healthy guinea pig cartilage samples on a 9.4 T Varian scanner.

2767.   Chemical Exchange Saturation Transfer effect from Phospho-creatine (PCr) and Adenosine-tri-phosphate (ATP) 
Mohammad Haris1, Kejia Cai1, Anup Singh1, Victor Babu KC1, Hari Hariharan1, and Ravinder Reddy1
1CMROI, Radiology, University of Pennsylvania, Philadelphia, Pennsylvania, United States

In the current study, we define the chemical exchange saturation transfer (CEST) effect from phospho-creatine (PCr) and adenosine-tri-phosphate (ATP). PCr showed CEST effect at ~2.5ppm while ATP at ~2.0ppm. CEST imaging at different concentration PCr and ATP solutions was performed and mapped. We are in process of implementing the current technique to investigate mapping of PCr and Cr in-vivo on human calf muscles both at 3T and 7T and monitor exercise related changes in these metabolites.

2768.   Quantitative modeling of in-vivo amide proton transfer measurements in the human brain indicates a dominant signal contribution from proteins with short T2 relaxation times 
Rachel Scheidegger1,2, Elena Vinogradov1,3, Weiying Dai1,3, and David C Alsop1,3
1Radiology, Beth Israel Deaconess Medical Center, Boston, MA, United States, 2Health Sciences and Technology, Harvard-MIT, Cambridge, MA, United States, 3Radiology, Harvard Medical School, Boston, MA, United States

Amide proton transfer imaging has the potential to measure pH in-vivo based on the amide exchange rate. Using an APT saturation scheme with simultaneous saturation at two frequencies we were able to remove MT contamination from the images, allowing for accurate quantification of amide exchange rate, relaxation time and concentration in the healthy human brain as a function of saturation power. The exchange rate was measured to be 45Hz, consistent with previous animal experiments. The transverse relaxation time of 2ms was shorter than previously assumed and may indicate the APT signal has a contribution from bound proteins.

2769.   Amide proton transfer imaging with continuous wave dual frequency saturation can detect the amide proton peak in the z-spectrum acquired at 3T 
Rachel Scheidegger1,2, Elena Vinogradov1,3, Weiying Dai1,3, and David C Alsop1,3
1Radiology, Beth Israel Deaconess Medical Center, Boston, MA, United States, 2Health Sciences and Technology, Harvard-MIT, Cambridge, MA, United States, 3Radiology, Harvard Medical School, Boston, MA, United States

We present a chemical exchange saturation transfer (CEST) imaging sequence with continuous wave saturation preparation relying on a 3-way subtraction between label frequency, control frequency, and simultaneous dual frequency RF irradiation to remove B0 inhomogeneity and intrinsic magnetization transfer (MT ) from in-vivo images. We demonstrate this approach yields amide proton transfer (APT) images free of susceptibility artifacts and MT asymmetry, without any additional B0 correction. This allows clear and robust measurement of the amide proton peak in the z-spectrum acquired at 3T. This new method may improve the feasibility of quantifying exchange rates in-vivo and measuring pH.

2770.   Optimization of pulsed-gagCEST at 3.0T 
Gopal Varma1, David C Alsop1, Robert E Lenkinski1, and Elena Vinogradov1
1Department of Radiology, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, United States

Chemical exchange saturation transfer from glycosaminoglycans (gagCEST) has shown potential in producing an endogenous contrast from cartilage, with applications involving study of cartilage degeneration. In moving to clinical 3.0T MR systems, a pulsed-CEST implementation has been shown to address SAR/hardware limits. Optimization is carried out by looking at different pulse shapes for implementation of the pulsed-CEST scheme. In particular, a gagCEST effect to differentiate normal bovine nasal cartilage from that degraded in trypsin is investigated and compared to results from a high-field scanner. The gauss shape is shown to produce the greatest gagCEST contrast, comparable to that at 4.7T.

2771.   Chemical Exchange Transfer Imaging of Creatine 
Anup Singh1, Mohammad Haris1, Kejia Cai1, Hari Hariharan1, and Ravinder Reddy1
1CMROI, Department of Radiology, University of Pennsylvania, Philadelphia, PA, United States

Creatine (Cr) plays an essential role in the storage and transmission of phosphate-bound energy. Using phantoms, it is demonstrated that CEST effect, between Cr (-NH2 protons) and bulk water protons (CrCEST), is highly sensitive to [Cr] concentration in physiological range and under physiological conditions at ultrahigh field(7T). In-vivo mapping of Cr using CrCEST is validated through modulation of Cr in rat brain tumor model. After intravenous injection of Cr solution, significant increase in CrCEST contrast as well comparable increase in Cr -CH2 resonance (in SVS) in is observed.

2772.   High Resolution Imaging of Myo-Insitol in Alzhemier’s Disease Pathology 
Mohammad Haris1, Anup Singh1, Kejia Cai1, Kavindra Nath2, Rachelle Berger1, Ari Borthakur1, Hari Hariharan1, and Ravinder Reddy1
1CMROI, Radiology, University of Pennsylvania, Philadelphia, Pennsylvania, United States, 2LMI, Radiology, University of Pennsylvania

The concentration of MI has been shown to change in Alzheimer’s disease (AD) pathology. Earlier, whole brain mapping of MI has been performed using chemical exchange saturation transfer (CEST) imaging. Here, we mapped the change in MI concentration, at high resolution, in transgenic AD mice brain (APP/PS1, n=3) using CEST technique. Compared to wild type mice (n=2), AD mice showed a significant increased MICEST contrast, which was also in good agreement with the changes in proton spectroscopy of MI. This method potentially provides a quantitative tool for measuring early molecular changes associated with the disease progression in AD.

2773.   CEST MRI of Human Liver at 3T 
Kejia Cai1, Anup Singh1, Kalli Grasley1, Mohammad Haris1, Damodar Reddy1, Hari Hariharan1, and Ravinder Reddy1
1CMROI, Department of Radiology, University of Pennsylvania, Philadelphia, PA, United States

Liver is a vital organ that regulates many key metabolites in the body. A number of metabolites with chemical exchange saturation transfer (CEST) effects have been shown to change in liver diseases. However, CEST MRI of liver in-vivo is vulnerable to motion artifacts. We have implemented a CEST FLASH (Fast Low Angle SHot) imaging sequence at 3T clinical scanner, with which artifact-free CEST images can be acquired in a single breath hold. CEST Z-spectral asymmetry profile from the liver has been characterized. CEST MRI of liver may open a new way to look in to liver diseases.

2774.   Identification of Endogenous Proteins Correlated with Amide Proton Transfer (APT) Imaging Contrast using Proteomic Analysis 
Kun Yan1, Zongming Fu2, Jennifer Van Eyk3, Silun Wang1, and Jinyuan Zhou1
1Radiology, Johns Hopkins University, Baltimore, MD, United States, 2Pediatrics, Johns Hopkins University, Baltimore, MD, United States, 3Bayview Proteomics Center, Johns Hopkins University, Baltimore, MD, United States

Amide proton transfer (APT) imaging as a specific type of chemical exchange¨Cdependent saturation transfer (CEST) MRI technique can distinguish brain tumor from normal tissue. To determine which protein(s) contribute to the APT signal, protein profiles of rat 9L tumor and normal brain tissues were compared by proteomic studies. Six significantly up-regulated cytosolic proteins were identified and supposed to correlate with APT hyper-intensity in tumor.

2775.   Keyhole Chemical Exchange Saturation Transfer 
Gopal Varma1, Robert E Lenkinski1, and Elena Vinogradov1
1Department of Radiology, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, United States

Contrast from chemical exchange saturation transfer (CEST) often requires multiple acquisitions. The keyhole technique is implemented that combines low-resolution data acquired at different saturation frequencies with a high-resolution reference image. CEST images reconstructed using the keyhole technique are compared with those obtained using standard acquisition at the same resolution and are shown to produce a comparable contrast for gagCEST. In particular, a keyhole image shows distinction of a cartilage sample from its surroundings that is not visible from the low-resolution data alone. Thus keyhole CEST provides a detectable contrast with significant reduction in acquisition time of the complete dataset.

2776.   3D Whole Brain Pulsed CEST Acquisition at 7T 
Craig K Jones1,2, Daniel Polders3, Jun Hua1, Hans Hoogduin4, He Zhu1, Jinyuan Zhou1,2, and Peter C M van Zijl1,2
1FM Kirby Center, Kennedy Krieger Institute, Baltimore, MD, United States, 2Department of Radiology and Radiological Sciences, Johns Hopkins Medical Institutes, Baltimore, MD, United States,3Department of Radiology, University Medical Center Utrecht, Utrecht, Netherlands, 4Brain Division, University Medical Center Utrecht, Utrecht, Netherlands

A 3D whole-brain pulsed CEST technique employing steady state saturation provide high resolution CEST images at 7T with limited to negligible interference of direct water or conventional magnetization transfer contrast. Amide proton transfer was calculated by fitting the acquired z-spectra to a Lorentzian, calculating the difference spectra and quantifying the mean difference between 3.2 and 3.8 ppm. Maps of the amide proton transfer was displayed from a normal volunteer.

2777.   Computational modeling and optimized detection of PARACEST contrast agents with Echo Planar Imaging 
Nevin McVicar1,2, Alex Li2, Adrienne Campbell3, Marty Klassen2, and Rob Bartha1,2
1Medical Biophysics, University of Western Ontario, London, ON, Canada, 2Centre for Functional and Metabolic Mapping, Robarts Research Institute, 3Division of Medicine and Institute of Child Health, UCL Centre for Advanced Biomedical Imaging, London, United Kingdom

A computational model of magnetic resonance imaging was developed that simulates the MR imaging of PARAmagnetic CEST (PARACEST) contrast agents using Echo-Planar Imaging (EPI) pulse sequences. The EPI PARACEST model simulated the imaging of a theoretical phantom containing 10 mM Eu3+-DOTAM-Gly-Phe. Using MATLAB, the model is used to strategically vary specific imaging parameters in order to optimize contrast to noise (CNR) efficiency. Several simulation experiments are presented with CNR efficiency displaying a clear dependence on the number of k-space segments, excitation flip angles, and CEST saturation pulse durations. Computational modeling provides insight into the dependence on specific imaging parameters.

2778.   Modeling MT Effect of Bound Water Pool and its use in Correction of CEST Contrast for MT Asymmetry 
Anup Singh1, Kejia Cai1, Mohammad Haris Haris1, Hari Hariharan1, and Ravinder Reddy1
1CMROI, Department of Radiology, University of Pennsylvania, Philadelphia, PA, United States

Magnetization transfer (MT) effect, exhibited by water protons bound to solid like macromolecules, measured using z-spectra technique of MRI, is used to explore macromolecular environment of tissue under consideration. MT asymmetry interferes with CEST contrast computation and usually results in underestimation or complete suppression of true CEST contrast. A method for modeling pure MT effect, using few frequency data of z-spectra, is presented. Modeled MT curve was used for removing MT asymmetry contribution in CEST contrast computation. Method was validated using simulations and in-vivo brain data. Presented method is simple, robust and insensitive to field in-homogeneities.

2779.   Two-pool compartmental modeling of balanced SSFP and CEST 
Kimberly Lara Desmond1, Sean Deoni2, Shannon Kolind2, and Greg J Stanisz1,3
1Medical Biophysics, University of Toronto, Toronto, ON, Canada, 2Oxford University, Oxford, United Kingdom, 3Imaging Research, Sunnybrook Health Sciences Centre, Toronto, ON, Canada

A quantitative two-pool compartmental model of CEST has been adapted to describe the signal observed with the balanced SSFP experiment. We examined the consequences of changing both RF phase and offset frequency on the characteristics of the asymmetry in the spectrum introduced by the off-resonance CEST component. Balanced SSFP sequences show promise for the measurement of CEST asymmetry and can be performed with a TR a hundred times less than that required for traditional continuous wave CEST experiments.

2780.   Optimized CEST Imaging of Intermediate to Fast Exchanging Agents in In-vivo Situations 
Anup Singh1, Hari Hariharan1, Kejia Cai1, Mohammad Haris Haris1, and Ravinder Reddy1
1CMROI, Department of Radiology, University of Pennsylvania, Philadelphia, PA, United States

In this study, optimization of saturation pulse amplitude and duration for in-vivo CEST imaging of intermediate to fast exchanging molecules is demonstrated using numerical simulations of Bloch-McConnell equations. High B1 amplitude and short duration of saturation pulse provides optimum CEST contrast with minimized contamination from slow exchanging agents and reduced SAR.

2781.   MRI of Glutamate Modulation in-vivo 
Kejia Cai1, Mohammad Haris1, Anup Singh1, Feliks Kogan1, Walter R.T. Witschey1, Prianka Waghray1, Joel H. Greenberg2, Hari Hariharan1, John A. Detre2, and Ravinder Reddy1
1CMROI, Department of Radiology, University of Pennsylvania, Philadelphia, PA, United States, 2Department of Neurology, University of Pennsylvania, Philadelphia, PA, United States

Glutamate (Glu) is the major neurotransmitter for fast excitatory synaptic transmission in central nervous system (CNS). Traditional magnetic resonance spectroscopy (MRS) for detection of Glu generally provides poor spatial and temporal resolution. This study demonstrates that Glu exhibits a pH-dependent chemical exchange saturation transfer (CEST) effect (GluCEST) in a concentration dependent manner. Glu modulation via intravenous injection in a rat brain tumor model resulted in clear elevation of GluCEST. Future studies using this approach may provide new insights into Glu function and demonstrate its potential as a biomarker for CNS disorders.

Traditional Posters : Pulse Sequences, Reconstruction & Analysis
Click on to view the abstract pdf and click on to view the pdf of the poster viewable in the poster hall.
Magnetization Transfer

Tuesday May 10th
Exhibition Hall  13:30 - 15:30

2782.   Drift in the magnetization transfer signal: effect on quantitative MT experiments 
Ives R Levesque1, Nikola Stikov2, G Bruce Pike2, and John M Pauly1
1Electrical Engineering, Stanford University, Stanford, CA, United States, 2Montreal Neurological Institute, McGill University, Montreal, Quebec, Canada

Drift in the signal of MT-weighted short-TR GRE sequences has been observed in quantitative MT experiments at two different sites. Signal decreases of 2-6% were observed over the course of data acquisitions in phantoms and in vivo. Drift appears to depend on the duration of the acquisition and on the MT pulse characteristics. The drift likely affects QMTI parameter estimates, in a manner dependent on the experimental method. We present our investigation into this effect, and discuss potential causes including B1 drift related to RF amplifier and/or coil performance, and temperature-related T1 increases.

2783.   Bound Pool Fraction and T1,free Quantification by Non-linear Parameter Identification of Composite Echoes 
Bernhard Neumayer1, and Rudolf Stollberger1
1Institute of Medical Engineering, Graz University of Technology, Graz, Steiermark, Austria

The method of phase acquisition of composite echoes (PACE) is used to additionally quantify BPF by applying a non-linear parameter identification and therefore gaining information on the separate components of the acquired signal. Extrapolation of the stimulated echo component for measurements with TM>100ms yields S0/(1+f) which in combination with S0 acquired using a minimum possible TM allows for quantification of the bound pool fraction BPF additionally to the T1 quantification provided by the phase information. The method was validated using BSA phantoms with concentrations ranging from 10 to 30% of BSA to water per weight and provided a linear relationship of BPF and BSA concentration.

2784.   Analysis of Magnetization Transfer Ratio Measurements at 3T Using Multiple-Acquisition Balanced SSFP 
Monika Gloor1, Klaus Scheffler1, and Oliver Bieri1
1Radiological Physics, University of Basel Hospital, Basel, Switzerland

Recently, it has been shown that fast magnetization transfer (MT) scans based on balanced steady-state free precession (bSSFP) with RF pulse modification achieve high SNR and high reproducibility at 1.5T. At 3T, combination of multiple, phase-cycled, bSSFP scans might be required to overcome limitations from off-resonance artifacts. In this study, different recombination methods for the generation of accurate magnetization transfer ratio (MTR) maps were analyzed. Maximum intensity projection removes off-resonance artifacts without significant MTR modification, whereas sum-of-squares and weighted-combination bSSFP yield reduced MTR values.

2785.   Preliminary investigation of the use of parallel RF Transmission in MTR measurement in the human cervical cord 
Rebecca Sara Samson1, Matthew Clemence2, Xavier Golay3, and Claudia A M Wheeler-Kingshott1
1Department of Neuroinflammation, UCL Institute of Neurology, London, England, United Kingdom, 2Philips Clinical Science Group, Philips Healthcare, Guildford, England, United Kingdom, 3UCL Institute of Neurology, United Kingdom

RF B1 transmit field non-uniformity, caused primarily by skin depth and dielectric resonance effects, is a large source of error in quantitative MR measurements made at 3.0T. We investigated the possibility that B1 errors could be reduced using dual transmission by measuring cervical cord MTR and B1 with and without dual transmission. Preliminary data acquired on three healthy subjects indicates that it may be possible to reduce inter-subject variation in whole cord MTR histogram peak locations using dual transmission at 3.0T. This could be an important consideration when designing future long-term clinical studies using quantitative MRI outcome measures.

2786.   Quantitative Magnetization Transfer Imaging of Human Brain at 7 Tesla 
Richard D Dortch1,2, Jay Moore2,3, Marcin Jankiewicz1,2, Adrienne N Dula1,2, Ke Li1,2, Daniel F Gochberg1,2, John C Gore1,2, and Seth A Smith1,2
1Radiology and Radiological Sciences, Vanderbilt University, Nashville, TN, United States, 2Institute of Imaging Science, Vanderbilt University, Nashville, TN, United States, 3Physics and Astronomy, Vanderbilt University, Nashville, TN, United States

Quantitative MT (qMT) imaging yields indices describing the interactions between free water and immobile macromolecular protons. Previous work at 1.5 and 3T has indicated that certain qMT indices may be sensitive to myelin content in white matter. Such studies may benefit from the increased SNR available at 7T; however, they are currently hampered by significant ΔB0 and B1+inhomogeneities. Therefore, we have developed a selective inversion recovery (SIR) qMT protocol at 7T that addresses these issues and here we report data acquired in healthy human brain. The resulting qMT parameters at 7T were in agreement with previously published values.

2787.   Magnetization Transfer Effects in Wideband SSFP 
Hung Phi Do1, Robert Marc Lebel2, and Krishna S Nayak2
1Department of Physics & Astronomy, University of Southern California, Los Angeles, California, United States, 2Department of Electrical Engineering, University of Southern California, Los Angeles, California, United States

Wideband SSFP is an alternating-TR sequence that suppresses much of the off-resonant banding artifacts associated with balanced SSFP. In this study, we demonstrate that wbSSFP is more sensitive to magnetization transfer (MT) than is bSSFP at all tip angles and RF pulse durations, making this sequence attractive for mapping the MT ratio or for imaging applications where strong MT weighting is desired.

2788.   Influence of Magnetisation Transfer on established T1 mapping methods 
Miriam Rabea Kubach1, Kaveh Vahedipour1, Tony Stoecker1, and N. Jon Shah1,2
1Forschungszentrum Juelich, Institute of Neuroscience and Medicine, Juelich, NRW, Germany, 2Department of Neurology, Faculty of Medicine, JARA, RWTH Aachen University, Aachen, Germany

Precise and accurate T1 mapping in the presence of MT requires modifications of the fitting model to account for the exchange. If those are neglected and the standard models are applied, an error in the T1 values is inevitable. Our work aims to quantify this error for different T1 mapping methods based on simulations of white and grey matter. The effects of SNR and MT on the precision and accuracy of T1 mapping was studied.

2789.   Exchange Resolved Measurements of Extra-cellular Volume in a Graded Muscle Edema Model 
Jack T Skinner1,2, Todd E Peterson2,3, and Mark D Does1,2
1Biomedical Engineering, Vanderbilt University, Nashville, TN, United States, 2Institute of Imaging Science, Vanderbilt University, Nashville, TN, United States, 3Radiology and Radiological Sciences, Vanderbilt University, Nashville, TN, United States

A simple, two-measurement method for inverting a two-pool model with water exchange is presented. T2 measurements of injured rat skeletal muscle were made before and after contrast agent injection, from which estimates of the intrinsic extra-cellular volume fraction were computed and found to correlate strongly (R2 = 0.71) with estimates derived from SPECT imaging of an extra-cellular radiotracer.

2790.   Correcting RF Inhomogeneities in Skeletal Muscle Magnetization Transfer Maps 
Christopher David James Sinclair1,2, Jasper M Morrow1, Michael G Hanna1, Mary M Reilly1, Tarek A Yousry1,2, Xavier Golay2, and John S Thornton1,2
1MRC Centre for Neuromuscular Diseases, UCL Institute of Neurology, London, United Kingdom, 2Department of Brain Repair and Rehabilitation, UCL Institute of Neurology, London, United Kingdom

We evaluate a scheme for correcting RF inhomogeneities in skeletal muscle magnetization transfer ratio (MTR) maps using B1 mapping data. A quantitative model of pulsed MT is used to demonstrate the applicability of the scheme to muscle and we perform automatic segmentation using T1 maps. We demonstrate experimentally that the correction scheme reduces the within- and between- subject variation of muscle MTR measures in 28 healthy volunteers and 23 patients with neuromuscular conditions. Reductions in instrumental inhomogeneities improve the potential of MTR as quantitative marker of disease, which has been shown previously to correlate with clinical disease status.

Traditional Posters : Pulse Sequences, Reconstruction & Analysis
Click on to view the abstract pdf and click on to view the pdf of the poster viewable in the poster hall.
Acquisition Strategies: From 3D to Spectroscopy

Wednesday May 11th
Exhibition Hall  13:30 - 15:30

2791.   Three Dimensional Imaging with Independent Slab Excitation and Encoding 
Amir Eissa1, and Alan H. Wilman1
1University of Alberta, Edmonton, Alberta, Canada

A new class of 3D Cartesian MRI is introduced whereby the voxel encoding orientation is independent of the imaging slab orientation. The excitation region can be chosen for the optimal region of interest, while the encoding direction can be arbitrarily oriented for ideal voxel alignment. The resulting aliased image can be easily unwrapped to produce standard image depiction. The method is particularly useful with anisotropic voxels, enabling voxel alignment to produce minimal blurring of oriented fine structures, while maintaining ideal volume placement. For phase susceptibility methods, the method enables oblique imaging with voxel orientation maintained along the main field direction.

2792.   2D RF Pulses with Rotating Read Out Direction for Increased FOV with Elevated Central SNR 
Andre de Oliveira1, Tobias K Block1, and Stephan Kannengiesser1
1Siemens AG, Erlangen, Germany

In MRI exams, focus is often given to a specific pathologic region (eg.: prostate, liver, brain) and in most cases the field of view (FOV) is centred at the region to be analysed. Here we propose a a new acquisition technique where 2D RF pulses to are employed and the RO direction is rotated for each individual average, allowing image acquisition with high central SNR (the same SNR as acquired with the actual protocols using 2D RF pulses) and increased FOV.

2793.   GESFIDE-PROPELLER for Simultaneous R2 and R2* Measurements in the Abdomen 
Ning Jin1, Yang Guo2, Jie Deng3, and Andrew C Larson1,4
1Departments of Radiology and Biomedical Engineering, Northwestern University, Chicago, IL, United States, 2Department of Radiology, Northwestern University, Chicago, IL, United States,3Department of Medical Imaging, Children's Memorial Hospital, Chicago, IL, United States, 4Robert H. Lurie Comprehensive Cancer Center, Chicago, IL, United States

Quantitative R2 and R2* maps are critical for a wide range of applications. The GESFIDE sequence can provide simultaneous R2 and R2* information. However, for abdominal imaging, application of this method can be particularly challenging due to respiratory motion. We developed a GESFIDE-PROPELLER approach for simultaneous R2 and R2* measurements in the abdomen and demonstrated that GESFIDE-PROPELLER can provide accurate R2 and R2* measurements while reducing respiratory motion artifacts.

2794.   Improved susceptibility weighted phase imaging for the assessment of brain iron deposition using a multi-echo sequence 
Guillaume Gilbert1,2, Geneviève Savard1, Céline Bard1, and Gilles Beaudoin1
1Department of Radiology, Centre Hospitalier de l'Université de Montréal, Montreal, QC, Canada, 2MR Clinical Science, Philips Healthcare, Cleveland, OH, United States

In this abstract, the use of a multi-echo gradient-echo sequence is investigated as a way to improve the contrast-to-noise ratio for susceptibility weighted phase imaging, which is often used to assess brain iron deposition. In comparison to the standard approach using a single-echo sequence, it is shown that the use of the proposed method allows for a significant reduction of the noise contribution in the reconstructed susceptibility weighted phase image, while preserving both contrast and acquisition time.

2795.   Multi-Directional High Moment Encoding in Phase Contrast MRI 
Nicholas Ryan Zwart1, and James Grant Pipe1
1Neuroimaging Research, Barrow Neurological Institute, Phoenix, Arizona, United States

Signal to noise ratio gains in low VENC phase contrast MRI are limited by the ability to successfully unalias phase measurements that fall outside the -180 to 180 degree interval. The ability to unalias phase measurements on a per pixel basis is limited by errors in the measurements due to noise and signal biased phase. The method presented in this work is the combination of a multiple low VENC acquisition strategy and reconstruction algorithm that produces time efficient SNR in phase contrast MRI.

2796.   Experimental Demonstration of nCPMG Realignment 
Patrick H Le Roux1, Graeme C McKinnon2, Yi-Fen Yen3, and Brice Fernandez4,5
1Applied Science Lab, GE Healhtcare, Palaiseau, France, 2Applied Science Lab, GE Healthcare, Waukesha, WI, United States, 3Applied Science Lab, GE Healthcare, Menlo-Park, CA, United States,4Applied Science Lab, GE Healhtcare, Nancy, France, 5IADI Lab, INSERM, Nancy, France

We present the first direct experimental proof of nCPMG capability to refocus the magnetization at the end of a long echo train. The nutation angle used is small enough (130°) to make a large difference between a sequence where realignment is used and a sequence where only stabilization is used. Also measuring the two transverse components insure the third one is also refocused, and that there is truly no initial phase dependence.

2797.   Two-Fold Phase Encoded SENSE Acceleration with a Single-Channel Coil 
Andre Jesmanowicz1, Andrew S. Nencka1, and James S. Hyde1
1Biophysics, Medical College of Wisconsin, Milwaukee, WI, United States

The addition to SENSE of spatial variation of the magnetization phase doubles the number of useful non-singular equations in the method. Phase encoded SENSE (P-SENSE) described here can be used without the need for spatial variation of coil sensitivities. An acceleration factor of up to 2 can be achieved in a single-coil MRI environment.

2798.   Spatially encoded ultrafast 2D SECSY in inhomogeneous fields 
Shuhui Cai1, Can Wu1, Zhiyong Zhang1, and Zhong Chen1
1Department of Physics, Fujian Key Laboratory of Plasma and Magnetic Resonance, Xiamen University, Xiamen, Fujian, China, People's Republic of

Ultrafast methods based on spatial encoding enable 2D and mD acquisition to be completed within a single scan and thus greatly shorten the experimental time. In this abstract, we proposed an ultrafast pulse sequence for acquiring high-resolution 2D SECSY spectrum in inhomogeneous fields. Absolute chemical shift and J-coupling information can be obtained. It works particularly well in the inhomogeneous fields where z-orientation inhomogeneities are dominant.

2799.   31P T1 measurement using ISIS with simultaneously measured spin-echo and stimulated-echo (ISIS-sSESTE) 
Xianfeng Shi1,2, Young-Hoon Sung1,3, SeongEun Kim2, Perry Renshaw1,3, and Eunkee Jeong2
1The Brain Institute, University of Utah, Salt Lake City, Utah, United States, 2Department of Radiology, University of Utah, Salt Lake City, Utah, United States, 3Department of Psychiatry, University of Utah, Salt Lake City, Utah, United States

Creatine Kinase (CK) is widely present in human brain tissues. It catalyzes the conversion between phosphocreatine (PCr) and adenosine diphosphate (ATP). Thus CK reaction rate constant (kf) is an indicator of the brain neuronal activity. A method using ISIS and a saturation recovery process provides accurate measurement of kf within 40 minute data acquisition, which is mostly spent to measure the apparent T1 relaxation time. The long acquisition time prevents its routine application. In this report, a new pulse sequence “ISIS with simultaneous spin echo and stimulated Echo (ISIS-sSESTE)” is developed to rapidly measure the 31P T1 relaxation time.

2800.   Optimized Chemical Shift Imaging for Sodium MRI of the Human Brain 
Patrick Michael Heiler1, Benedikt Rieger1, Philipp Krämer1, Simon Konstandin1, and Lothar Rudi Schad1
1Computer Assisted Clinical Medicine, Heidelberg University, Mannheim, Germany

During the long phase encoding of sodium CSI measurements, due to hardware constraints on common whole-body scanners, signal significantly decays meanwhile no data acquisition is possible. Thus, the presented work describes a 3D chemical shift imaging (CSI) sequence with individually minimized phase encoding durations for sodium MRI. In an additional experiment, CSI data is acquired with a Hanning weighted pre-filter and both approaches are compared to the radial projection imaging technique.

2801.   Echo Planar based J Resolved and Correlated Spectroscopic Imaging of Human Prostate Using External Coil 
Rajakumar Nagarajan1, Jonathan Furuyama1, Daniel Margolis1, Steven Raman1, Manoj Kumar Sarma1, and Michael Albert Thomas1
1Radiological Sciences, University of California Los Angeles, Los Angeles, California, United States

Prostate cancer (PCa) is associated with lower levels of citrate (Cit) and higher levels of choline (Cho) than those in benign prostatic hyperplasia or healthy prostate tissues. Due to the overlap of Cho with creatine resonances, it is difficult to measure spermine resonances in the prostate clearly by single-and multi-voxel based one-dimensional spectroscopic imaging. We have implemented and evaluated the novel echo-planar imaging based four dimensional (4D) MRSI sequences (J-resolved spectroscopic imaging (EP-JRESI) and correlated spectroscopic imaging (EP-COSI)) in the prostate using 3T MRI scanner. The pilot findings in six healthy males using the external phased-array matrix will be presented.