Relaxometry-Technical Developments

Monday 1 June 2015

Thomas Christen¹, Wendy W Ni¹, Samantha Holdsworth¹, Murat Aksoy¹, Roland Bammer¹, Michael Moseley¹, and Greg Zaharchuk^1
1Department of Radiology, Stanford University, Stanford, California, United States

In this work, we studied the effects of B0 inhomogeneities on MR fingerprinting acquisitions when short TEs, multiple TEs, long TRs, or gradient spoilers are used during a SSFP acquisition, and when virtual linear shim gradients are added into the fingerprinting dictionary.

Mahesh Bharath Keerthivasan¹, Ali Bilgin^1,2, Diego R Martin², and Maria I Altbach^2
1Electrical and Computer Engineering, University of Arizona, Tucson, Arizona, United States, ²Medical Imaging, University of Arizona, Tucson, Arizona, United States

T2-weighted imaging and T2 mapping play an important role in the diagnosis of pathologies and in the quantitative characterization of tissue. 3D imaging and 3D T2 mapping are useful in applications where the anatomical structure can be better viewed by reformatting an isotropic voxel. In this work, we present a technique for isotropic T2 mapping that uses data collected using a radial 3D Fast Spin Echo pulse sequence and a reconstruction technique that generates T2 maps from highly undersampled data.

Damien Nguyen¹ and Oliver Bieri^1
1Radiological Physics, Dep. of Radiology, University of Basel Hospital, Basel, Switzerland

In this work, we present a new approach for fast motion-insensitive relaxometry based on a N-point discrete Fourier analysis of a series of N balanced steady-state free precession (bSSFP) scans with shifted frequency response profiles. Simulations, as well as phantom and human experiments are performed to validate the new method.

Wenbo Li^1,2, Ksenija Grgac^1,2, Alan Huang^1,3, Qin Qin^1,2, Nirbhay Yadav^1,2, and Peter Van Zijl^1,2
1Department of Radiology, Johns Hopkins University School of Medicine, Balitmore, MD, United States, ²F.M. Kirby Research Center for Functional Brain Imaging, Kennedy Krieger Institute, Baltimore, MD, United States, ³Current Address: Philips Healthcare, Best, Netherlands

Here we present a theory for the water proton T1 of blood that takes into account the effect of hematocrit, oxygenation and hemoglobin concentration, and validate it on extensive T1 data acquired from whole blood and lysed blood at 3T, 7T, 9.4T, 11.7T and 16.4T. The theory involves the contributions of different mechanisms (paramagnetic effect, water tumbling hindrance) to blood T1 and can be used to predict blood T1 at multiple fields. As such it can provide reference values for perfusion-based quantification experiments such as ASL (Arterial Spin Labeling) and VASO (Vascular Space Occupancy).

Jean-David Jutras¹, Keith Wachowicz^1,2, and Nicola DeZanche^1,2
1Oncology, University of Alberta, Edmonton, AB, Canada, ²Medical Physics, Cross Cancer Institute, Edmonton, AB, Canada

DESPOT2 was proposed for fast, high-resolution, T₂ mapping but it is susceptible to banding artifacts induced by B₀ inhomogeneity, especially at 3T and higher fields. We have generalized the DESPOT2 technique by deriving an analytical expression for T₂ using two or more phase-cycled bSSFP images at varying flip angles. Our method can also be used to map B₀ inhomogeneity.

James A. Rioux¹, Ives R. Levesque^1,2, and Brian K. Rutt^1
1Radiology, Stanford University, Stanford, CA, United States, ²Medical Physics Unit, and Research Institute of the McGill University Health Centre, McGill University, Montreal, QC, Canada

White matter exhibits biexponential relaxation, which is neglected in T₁ measurements at low field strengths, but which is a significant source of variation between reported T₁ values at high field. If the effect of the shorter T₁ component can be removed, such that only the longer component is measured, the resulting T₁ value will be more reproducible across sites and protocols. We characterized biexponential relaxation in white matter at 7T, finding T₁ values of 1349ms and 57ms, with 11% of signal from the short component. Based on these values we recommend using minimum inversion times of 150-200ms in inversion-recovery-based sequences.

John Thomas Spear^1,2, Xiaoyong Zhang^2,3, and John Gore^2,3
1Physics and Astronomy, Vanderbilt University, Nashville, TENNESSEE, United States, ²Vanderbilt University Institute of Imaging Science, Vanderbilt University, Nashville, TENNESSEE, United States, ³Department of Radiology, Vanderbilt University, Nashville, TENNESSEE, United States

R₁ dispersion measurements are used to quantify sub-voxel microstructure in rat liver. Liver was chosen due to its relatively homogeneous makeup and for the fact that a Gd-DTPA injection will stay extracellular. Dispersion in R₁ at low locking fields has been shown to be caused by diffusion through internal susceptibility gradients. The inflection of the corresponding dispersion provides insight into the dimensions of the structures that perturb the local field. The data presented suggest the dispersion originates from the microvasculature and not the individual cells. The preliminary results are promising but further validation is required.

Michael John Knight¹, Bryony Wood¹, Elizabeth Coulthard², and Risto Kauppinen^1
1School of experimental psychology, University of Bristol, Bristol, Avon, United Kingdom, ²Southmead Hospital, University of Bristol, Bristol, Avon, United Kingdom

Magnetic resonance relaxation parameters are highly sensitive to the microstructure and biochemistry of tissues. We show here that spin-echo T2 can be related to fiber orientation and potentially myelin-associated water fraction through measurements of its anisotropy in the human brain. We also propose a novel model for nuclear spin relaxation anisotropy which we validate with data from human volunteers. Such measurements may be used to provide a more detailed and objective assessment of tissue state than is apparent by visual inspection of images.

Chu-Yu Lee^1,2, Xingju Nie^1,2, Jens H Jensen^1,2, Vitria Adisetiyo^1,2, Qingwei Liu³, and Joseph A Helpern^1,2
1Department of Radiology and Radiology Science, Medical University of South Carolina, Charleston, SC, United States, ²Center for Biomedical Imaging, Medical University of South Carolina, Charleston, SC, United States, ³Neuroimaging research, Barrow Neurological Institute, Phoenix, AZ, United States

In biological tissues, the presence of iron-rich cells, deoxygenated red blood cells or a paramagnetic agent generates micron-scale variations of magnetic susceptibility, resulting in microscopic magnetic field inhomogeneities (μMFI). Therefore, it is possible to characterize in vivo tissue properties through quantifying the μMFI. The relaxation rates R2, R2*, and R2′ have been previously used to quantify the relaxation due to μMFI. An alternative approach is magnetic field correlation (MFC) imaging, where the measured MFC is closely linked to the μMFI. MFC has been shown to effectively reflect iron depositions in the brain during normal aging and disease processes. A prior study compared MFC, R2, and R2* for cell suspensions with different Gadolinium (Gd) contrast agent concentrations. However, the distinction between these measures has not been investigated with structure-induced variable μMFI. In this work, we investigate how MFC, R2, R2*, R2′ change in phantoms with distinct μMFI properties.

Nana K. Owusu¹, Casey P. Johnson², William R. Kearney², John A. Wemmie^3,4, and Vincent A. Magnotta^2
1Biomedical Engineering, University of Iowa, Iowa City, IA, United States, ²Radiology, University of Iowa, Iowa City, IA, United States, ³Psychiatry, University of Iowa, Iowa City, IA, United States, ⁴Veterans Affairs Medical Center, Iowa City, IA, United States

The sensitivity of T1ρ to pH and glucose concentration is of great interest for imaging a variety of brain diseases. However, little work has been done on clinical imaging systems (≤3T) to assess the sensitivity and improve the specificity of T1ρ to these metabolic factors. To address this, we conducted pH and glucose phantom experiments at 3T. We found that T1ρ, compared to T2, provides superior sensitivity to pH but not glucose. These findings suggest that T1ρ may be sensitive to brain pH dynamics at 3T and that multi-parametric mapping of T1ρ and T2 may improve acquisition specificity to pH.

Chu-Yu Lee^1,2 and Jens H Jensen^1,2
1Department of Radiology and Radiology Science, Medical University of South Carolina, Charleston, SC, United States, ²Center for Biomedical Imaging, Medical University of South Carolina, Charleston, SC, United States

In patients with iron overload, the CPMG measurement in liver exhibits a non-monoexponential decay. By using a theoretical model, one can decompose the decay into two major components associated with the two forms of endogenous tissue iron: ferritin and hemosiderin. A separate quantification may be useful in assessing iron storage and chelation therapy, because ferritin iron may more directly reflect iron toxicity. However, the validity of the quantitative model has not been previously studied with numerical simulations of in vivo iron deposits with heterogeneous sizes and spatial distributions. In this work, we investigate the model’s performance in realistic tissue geometry by simulating the MR relaxation from iron spheres with statistically distributed sphere sizes. The simulated iron concentrations of ferritin and hemosiderin were varied over a clinically relevant range. As a reference, conventional bi-exponential fits to the signals were also calculated.

Casey P. Johnson¹ and Vincent A. Magnotta^1
1Radiology, University of Iowa, Iowa City, IA, United States

T1ρ mapping has found application for a variety of brain diseases. However, to accurately quantify T1ρ in the brain, particularly in difficult regions near air-tissue interfaces such as portions the frontal lobe, new spin-lock prep pulses are needed. This work simulates the T1ρ quantification accuracy of a number of spin-lock prep pulses, including those with adiabatic excitation and refocusing pulses, over an expected range of B1 and B0 field inhomogeneities calculated from 24 in vivo whole-brain scans. It is shown that use of adiabatic excitation pulses improves T1ρ accuracy in regions of the brain susceptible to B1 and B0 offsets.

Taehwa Hong¹, Min-Oh Kim¹, Dongyeob Han¹, Dosik Hwang¹, and Dong-Hyun Kim^1
1Electrical & Electronic Engineering, Yonsei University, Seodamun-gu, Seoul, Korea

Many of quantification technique of multiple properties of tissues can be hampered by B1 inhomogeneity at high field strengths. MRF combined with AFI technique is proposed to achieve simultaneous B1+ mapping and B1+ map compensated MRF

Jun Chen¹, Michael Carl², Hongda Shao¹, Qun He¹, Eric Chang^1,3, Christine B Chung^1,3, Graeme M Bydder¹, and Jiang Du^1
1Radiology, University of California, San Diego, CA, United States, ²GE Healthcare, San Diego, CA, United States, ³Department of Radiology, VA San Diego Healthcare System, San Diego, CA, United States

Numerous studies have revealed the existence of bound and free water in the bone as well as their different contributions to the mechanical properties of cortical bone. In this study we firstly compared a series of Ultrashort echo time (UTE) and adiabatic inversion recovery prepared UTE (IR-UTE) techniques in measurements of T2* and T1 of bound and pore water in cortical bone , then introduced 3D Cones sequences to quantify T2* and T1 values of different water components in vivo using a clinical whole-body 3T scanner. Consistent values of bound and pore water T2*s, their fractions as well as T1 could be observed using 2D UTE and 3D Cones sequences both in bovine bone and healthy volunteers, suggesting 3D Cones sequences can effectively measure both T2* and T1 of bound and pore water in cortical bone.

Rahel Heule¹, Carl Ganter², and Oliver Bieri^1
1Division of Radiological Physics, Department of Radiology, University of Basel Hospital, Basel, Switzerland, ²Department of Radiology, Klinikum rechts der Isar, Technische Universität München, Munich, Germany

Many T₁ mapping methods have been developed, however, they deliver substantially varying quantification results for the same tissues and at equal field strength. The variable flip angle (VFA) method based on radiofrequency spoiled gradient echo imaging provides 3D MR data sets and allows for volumetric T₁mapping in clinically feasible times. In this work, a new postprocessing approach is proposed to overcome the T₂-related bias originating from incomplete spoiling in conventional VFA T₁ quantification. Excellent agreement with inversion recovery reference measurements was obtained in vivo for human brain tissues demonstrating the potential of the new method to provide accurate whole-brain T₁ quantification at high-resolution.

Kecheng Liu¹, Dan Ma², Tiejun Zhao¹, and Mark Griswold^2
1Siemens Medical Solutions USA, Inc., Malvern, PA, United States, ²Case Western Reserved University, Cleveland, OH, United States

T2 quantification has been used for long time in clinical routine diagnosis, which relies objectively on an absolute quantitative T2-map rather than subjectively on gray-scaled images. Therefore, the reliability of quantified values should be ensured for clinical diagnosis. Currently the widely used multiple contrast spin echo acquisition mainly suffers from non-ideal RF pulse profiles, resulting in protocol dependent quantifications. Following our previous work, this study presents a true CPMG prepared T2-mapping method, which minimizes the impacts of the RF pulse profile, yielding more accurate T2-maps.

Charles Robert Castets¹, Emeline Julie Ribot¹, Aurélien Julien Trotier¹, William Lefrançois¹, Jean-Michel Franconi¹, and Sylvain Miraux^1
1RMSB - UMR5536, CNRS - Université de Bordeaux, Bordeaux, Aquitaine, France

In this study, a fast 3D T1 mapping method using stack-of-spirals sampling scheme and continuous radio-frequency pulse excitation at high magnetic field is presented. This sequence with a new fitting method has been validated to detect variations of Manganese concentrations in vivo and in vitro on both healthy and pathological mice models. Combining this method with a manganese injection allowed to accurately characterize tissue damages induced by a myocardial infarction.

Ken-Pin Hwang^1,2, Kevin King³, Peng Lai³, Wolfgang Stefan², Christopher McClellan², Ersin Bayram¹, and Ajit Shankaranarayanan^3
1Global MR Applications and Workflow, GE Healthcare, Houston, TX, United States, ²Department of Imaging Physics, The University of Texas MD Anderson Cancer Center, Houston, TX, United States, ³Global MR Applications and Workflow, GE Healthcare, Waukesha, WI, United States

Qmap is a multi-parameter mapping sequence that can acquire images for simultaneous T1, T2, and PD mapping within clinically acceptable acquisition times. Currently this is achieved with standard parallel imaging techniques applied to the individual slices. Potential exists to accelerate the acquisition further, especially if correlated information among the images is utilized. Compressed sensing and kat-ARC acceleration techniques were applied to a fully sampled Qmap acquisition. Synthetic images were produced with little error relative to the fully sampled acquisition, suggesting that synthesis of relative contrast images may be robust to the acceleration methods applied to the acquisition.

Cihat Eldeniz¹, Jürgen Finsterbusch², Weili Lin¹, and Hongyu An^1
1University of North Carolina at Chapel Hill, Chapel Hill, NC, United States, ²Universitätsklinikum Hamburg-Eppendorf, Hamburg, Germany

T1 mapping is beneficial in many respects. However, as with all other time-series acquisitions, motion is a significant threat. Our sequence is specifically designed to be robust to motion. GRAPPA calibration scans are performed multiple times for artifact-free image reconstruction and sub-voxel spin tracking is applied to correct for motion-related spin history distortions. The in vivo results show the promise of the approach.

M. Dylan Tisdall^1,2 and André J. W. van der Kouwe^2,3
1Martinos Center for Biomedical Imaging, Massachusetts General Hospital, Charlestown, Massachusetts, United States, ²Radiology, Harvard Medical School, Boston, MA, United States, ³Martinos Center for Biomedical Imaging, Massachusetts General Hospital, Charlestown, MA, United States

We demonstrate the addition of variable length echo trains to AFI to improve the SNR/time of the scan, reduce distortions due to B₀ inhomogeneity, and allow for accurate T₂^* mapping. We also introduce a novel, highly efficient algorithm for finding the maximum likelihood estimate of T1 and flip angle scaling error from the multi-echo data.

Ruiliang Bai^1,2, Alexander Cloninger³, Wojciech Czaja⁴, and Peter J. Basser^1
1Section on Tissue Biophysics and Biomimetics, National Institutes of Health, Bethesda, Maryland, United States, ²Biophysics Program, University of Maryland, College Park, Marland, United States, ³Applied Mathematics Program, Yale University, New Haven, Connecticut, United States, ⁴Department of Mathematics, University of Maryland, College Park, Maryland, United States

The power of 2D relaxation spectrum NMR and MRI to characterize complex water dynamics (e.g., compartmental exchange) in biology and other disciplines has been demonstrated in recent years. However, the large amount of data and long MR acquisition times required for conventional 2D MR relaxometry limits its applicability for in vivo preclinical and clinical MRI. We present a new MR pipeline that incorporates compressed sensing (CS) as a means to vastly reduce the amount of relaxation data needed for material and tissue characterization without compromising data quality. This framework is validated using synthetic data, with NMR data acquired in a well-characterized urea-phantom, and on fixed porcine spinal cord tissue.

Hasan Celik¹, Ariel Hafftka², Alexander Cloninger³, Wojciech Czaja², and Richard G. Spencer^1
1National Institute on Aging, National Institutes of Health, Baltimore, MD, United States, ²Department of Mathematics, University of Maryland, College Park, MD, United States, ³Applied Mathematics Program, Yale University, New Haven, CT, United States

Use of compressed sensing (CS) can significantly decrease acquisition times for multidimensional NMR relaxometry and related experiments. . In the present work, we apply the CS approach of Cloninger et al. for the first time to two experimental systems, olive oil and bovine patella articular cartilage. Further, we present an extension of the algorithm from 2D to 3D, with further experimental validation using the olive oil phantom.

Debra E. Horng^1,2, Diego Hernando¹, and Scott B. Reeder^1,2
1Radiology, University of Wisconsin-Madison, Madison, WI, United States, ²Medical Physics, University of Wisconsin-Madison, Madison, WI, United States

R2*-based measurement of liver iron is appealing due to its rapid acquisition and its linear relationship with iron concentration. Estimation techniques generally fall into two main categories: "fit-first" (FF), where an R2* map is generated at each voxel, followed by region-of-interest (ROI) averaging, and "average-first" (AF) techniques, where the signal is averaged over an ROI, followed by fitting an R2* estimate. We examine the relative performance of 6 techniques in simulations and phantoms over 50s^-1 to 2000s^-1. The combination of high R2* and low SNR leads to poor performance in the phantom, likely requiring higher SNR/shorter TEs for accurate R2* relaxometry.

Rui Pedro A. G. Teixeira^1,2, Shaihan J. Malik^1,2, and Joseph V. Hajnal^1,2
1Center for the Developing Brain, King's College London, London, United Kingdom, ²Division of Imaging Sciences and Biomedical Engineering, King's College London, London, United Kingdom

The DESPOT method for measuring T1 and T2 provides efficient high resolution mapping that is widely used in brain studies, but is regarded as at risk of overestimating relaxation time values compared to more conventional slower methods. We hypothesized that this could be caused by imperfect spoiling of the T1 weighted spoiled gradient echo sequences used. We explored a dictionary based fitting approach in which the SPGR signal is explicitly modeled using an extended phase graph. Results suggest that while the method allows fitting with lower residuals, resulting parameters values are not significantly changed.

Relaxometry Applications

Monday 1 June 2015

Matthew Tarasek¹, Oguz Akin², Jeannette Christine Roberts³, Tom Foo¹, and Desmond T.B. Yeo^1
1MRI, GE Global Research, Niskayuna, NY, United States, ²Radiology, MSKCC, New York, NY, United States, ³Imaging & Physiology Lab, GE Global Research, Niskayuna, NY, United States

MR imaging provides excellent spatial resolution and anatomical soft tissue contrast, yet there are still limitations in detecting and delineating early-stage cancer lesions when they are curable. In hopes of extending ideas for a multiparametric, quantitative MRI data set, we evaluated a unique approach for MR contrast by utilizing the thermal responses of heat-sensitive MR parameters. We look for the accuracy and repeatability in measuring these parameters, and evaluate if heat-induced contrast mechanisms have the potential to add information to conventional MR imaging contrast types for better identification and characterization of in vivo tumors.

Daniel Barazany^1,2, Ella Striem-Amit³, Shani Ben Amitay¹, Amir Amedi³, and Yaniv Assaf^1
1Neurobiology, Tel Aviv University, Tel Aviv, Israel, ²CUBRIC School of Psychology, Cardiff University, Cardiff, United Kingdom, ³Medical Neurobiology, The Hebrew University of Jerusalem, Jerusalem, Israel

IR-MRI provides enhanced image intensity in the human cortex that enables its subdivision to laminar shaped clusters. These clusters reflect cytoarchitectonic organization in the cortex, which could be further used for visualizing and quantifying changes due to brain impairment. In this study, IR-MRI analysis framework was employed on congenitally blind and sighted control subjects focusing on the visual cortices (V1 and V2). We explored the clusters composition on each region, aiming to expose microstructural differences originating probably from lack of input to the visual cortex as occurred in blindness.

Lina Avancini Colucci¹, Matthew Li¹, and Michael J. Cima^2
1Health Sciences and Technology (HST), MIT, Cambridge, Massachusetts, United States, ²Materials Science and Engineering, MIT, Cambridge, Massachusetts, United States

This study investigated the use of whole-body NMR as a quantitative, sensitive, non-invasive technique of measuring fluid accumulation in the IP space of mice. Mice were given an IP injection of different saline volumes and their whole body T2 relaxation time was measured. The magnetization decay data was put through an inverse Laplace transform to generate a relaxogram. The saline injections resulted in an amplitude increase in the relaxogram’s long peak. The increase of the area under the peak was proportional to the volume of injected fluid. Ascitic fluid, similarly, has a long relaxation time compared to other body tissues.

Ulrike Noeth¹, Elke Hattingen², Oliver Baehr³, Julia Tichy³, and Ralf Deichmann^1
1Brain Imaging Center (BIC), Goethe University Frankfurt/Main, Frankfurt/Main, Germany, ²Institute of Neuroradiology, University Hospital Frankfurt/Main, Germany, ³Dr Senckenberg Institute of Neurooncology, Goethe University Frankfurt/Main, Germany

Synthetic MP-RAGE anatomies were calculated from quantitative parameter maps (T1, PD, and RF coil bias) and the influence of these parameters on MP-RAGE image contrasts is discussed. Acquired and synthetic data are analyzed for SNR, CNR, and optical contrasts. In contrast to conventional anatomies with mixed contrasts, purely T1-weighted synthetic anatomies show significantly higher optical contrasts and improved visibility of tumor and edema in patients with high grade glioma.

Alexia Daoust¹, Stephen Dodd¹, Govind Nair¹, Steven Jacobson¹, Daniel S. Reich¹, and Alan Koretsky^1
1NINDS, LFMI, NIH, Bethesda, MD, United States

There continues to be interest in using changes in CSF properties to image neurodegenerative diseases. To optimize MRI sequences that enable segmentation of CSF from tissue, we characterized the CSF relaxometric properties at various field strengths in vivo and in vitro. We have shown an important difference of in vitro CSF T2 vs saline T2 that is not explained with CSF metals. Proteins or lipids could influence CSF T2. Our in vitro results suggest that in vivo T2 value at high field is incorrect and that low field is more optimal to quantify CSF relaxivity in vivo.

Kristin Toy¹, Eamon Doyle^1,2, Thomas Coates³, and John C Wood^1,2
1Cardiology, Children's Hospital of Los Angeles, Los Angeles, CA, United States, ²Biomedical Engineering, University of Southern California, Los Angeles, CA, United States, ³Hematology-Oncology, Children's Hospital of Los Angeles, Los Angeles, CA, United States

We validate a previously derived model for relaxivity enhancement due to iron in the liver, spleen, kidney, pancreas, and heart.

Kristin Toy¹, Eamon Doyle², Thomas Coates³, and John C Wood^1
1Cardiology, Children's Hospital of Los Angeles, Los Angeles, CA, United States, ²Biomedical Engineering, University of Southern California, Los Angeles, CA, United States, ³Hematology-Oncology, Children's Hospital of Los Angeles, Los Angeles, CA, United States

We propose R1 as a mechanism for determining liver iron load at 3T due to increased sensitivity relative to 1.5 T.

Lech Wiktor Skorski¹, Dorota Wierzuchowska², and Barbara Blicharska^1
1Radiospectroscopy, Jagiellonian University, Krakow, Malopolskie, Poland, ²Pedagogical University, Krakow, Malopolskie, Poland

The new results obtained by NMR relaxation method, namely the time evolution of relaxation time in blood serum, after initiation of oxidation process by addition of H2O2 are presented. Hydrogen peroxide is one of the strongest reactive oxygen species and when added to aqueous solutions it causes changes of the water proton relaxation times. Just after H2O2 addition to protein solution samples the relaxation time starts to short and after approximately 25 minutes stabilizes. In serum, after reaching the minimum value T1 instead of stabilization starts to regrow. It means that this regrow might be a consequence of the action of antioxidants which are known to be present in blood serum. We hope NMR relaxation measurements may be useful in diagnosis of some diseases, specially with free radical background.

Erik M. Akkerman¹, Jurgen H. Runge¹, Marian A. Troelstra¹, Aart J. Nederveen¹, and Jaap Stoker^1
1Radiology, Academic Medical Centre, Amsterdam, North Holland, Netherlands

From 01-2008 until 12-2013 we performed MR-measurements of liver iron content (LIC) in 95 patients, using a standardized protocol, including R2*-measurement, and LIC-assessment following Gandon's method. To optimize B0-homogeneity, we measured only three slices simultaneously, with shimming performed on a manually defined region, excluding arms and air. The same liver ROIs were used to analyze both methods. The resulting relation between R2* and estimated LIC showed a low error variation and was significantly non-linear. It could be well described by a quadratic function. Previous publications only reported linear relations, probably due to a higher accuracy in our measurements.

Roberto Salvati^1,2, Eric Hitti^1,2, Herve Saint-Jalmes^1,2, Robert Mulkern^3,4, and Giulio Gambarota^1,2
1Université de Rennes 1, LTSI, Rennes, France, ²INSERM, UMR 1099, Rennes, France, ³Department of Radiology, Boston Children's, Boston, MA, United States,⁴Harvard Medical School, Boston, MA, United States

Corn starch with water creates a white liquid which has generated particular interest in several groups. The relaxation times values are close to the ones we could have in vivo. A tube filled with water and a concentration of 66% in volume of corn starch was prepared. Two additional tubes, filled with agar gel and agar gel were also added to the phantom as reference. Corn starch solution with water displays interesting MR relaxation characteristics, with low T2 and T2* values and low T2/T1 ratio. As such, it could serve as a valuable phantom for testing short T2 and T2*.

Thomas Frenzel¹, Pavol Szomolanyi², Iris Noebauer-Huhmann², Martin Rohrer¹, Gregor Jost¹, and Siegfried Trattnig^2
1Bayer Healthcare, Berlin, Germany, ²High Field MR Centre, Department of Biomedical Imaging and Image-Guided Therapy, Medical University of Vienna / Vienna General Hospital, Vienna, Austria

The concentration dependent change of the signal intensity (SI) of commercially available contrast media (CM) were investigated in-vitro in human plasma at 37°C at 3T and 7T, using spin- and gradient echo sequences. A nearly linear correlation between Gd-concentration and SI was observed only for very low concentrations (<1-3mM). At higher concentrations stable, increasing or decreasing SIs were observed. The dominant parameters influencing the SI were the pulse sequence, the field strength, the protein binding of the CA as well as their relaxivities, and their in-vivo concentrations. These results should be considered for (semi-) quantitative image evaluation.

Kristin Toy¹, Eamon Doyle², Thomas Coates³, and John C Wood^1
1Cardiology, Children's Hospital of Los Angeles, Los Angeles, CA, United States, ²Biomedical Engineering, University of Southern California, Los Angeles, CA, United States, ³Hematology-Oncology, Children's Hospital of Los Angeles, Los Angeles, CA, United States

We investigate the relationship between R1 and liver iron load at 1.5T to assess the usefulness of R1 as a diagnostic tool and to provide in vivo validation of in vitro and animal model results.

Ping Wang¹, Isaac V Manzanera Esteve¹, Charles Nockowski², and John C Gore^1
1Vanderbilt University Institute of Imaging Science, Nashville, TN, United States, ²Philips Healthcare Technical Support at Vanderbilt, Nashville, TN, United States

The effects of T1 on estimates of sodium content in human muscle have been evaluated. Our data confirm that sodium levels estimated by comparing MRI signals to calibration phantoms are inaccurate unless the relaxation times of the tissue and phantom match, or unless longer TR values are used so that both tissue and phantom are fully relaxed. When simple solutions of sodium are used as references standards, if TR is greater than 130ms, the relative error in the reported tissue sodium content is reduced to < 5%.

Gunther Helms^1,2, Christina Schlumbohm³, Enrique Garea-Rodriguez^4,5, and Eberhard Fuchs^5
1Medical Radiation Physics, Lund University, Lund, Scania, Sweden, ²Cognitive Neurology, Göttingen University Medical Center, Göttingen, Lower Saxony, Germany, ³Encepharm Inc., Göttingen, Lower Saxony, Germany, ⁴Neuroanatomy, Albert-Ludwigs-University Freiburg, Freiburg, Badenia, Germany, ⁵Clinical Neurobiology Group, German Primate Center, Göttingen, Lower Saxony, Germany

Common marmoset monkeys (Callithrix jacchus) are increasingly used in models of neurodegenerative diseases. Serial measurements of gadolinium excretion were performed using the dual flip angle method to map R1 on three animals in the wrist coil of a 3T whole-body system. R1 values were analyzed in a ROI in the straight sinus and fitted with a single compartment model. Slow injection of 0.3 mmol Gadovist per kg bodyweight (2 min duration) resulted in faster equilibration and excretion, as well as smaller distribution volumes than fast injection (15 sec). Slow injection and an equilibration delay of 15 minutes are recommended.

Lech Wiktor Skorski¹ and Barbara Blicharska^2
1Radiospectroscopy, Jagiellonian University, Krakow, Malopolskie, Poland, ²Radiospectroscopy, Jagiellonian University, Malopolskie, Poland

It is well known that the presence of free paramagnetic ions shortens the NMR relaxation times of blood serum. Addition of the selective chelating agent (i.e.: D-penicillamine (D-pen) to serum causes the formation of stable and water soluble complexes of free ions and this nulls the effect of shortening of relaxation times of solution. After chelation one can easily measure the prolongation of relaxation times and this fact can be used as evidence of presence of free ions in blood serum. Obtained results confirm that the chelate action of Riboflavin (Vit.B2) is very similar to action of D-pen, but riboflavin as a natural product, have much smaller side effects and will be better tolerably for patients as D-pen.

Steven Reynolds¹, Peter R Laity², Ben Curie¹, Chris Holland², and Martyn N Paley^1
1Academic Unit of Radiology, University of Sheffield, Sheffield, South Yorkshire, United Kingdom, ²Department Materials Science and Engineering, University of Sheffield, Sheffield, South Yorkshire, United Kingdom

Silkworms form fibres from a proteinaceous mixture, silk dope, in two glands. The posterior section of the gland secretes proteins (fibroin), which are stored in the mid-section before being spun through a tapered silk duct. Rheology of extracted silk dope show chemical and mechanical changes during spinning. We determine the T1/T2 per voxel in live silkworms along the gland. The T1 and T2 varied smoothly along the gland, with T1 significantly different between the tapered and middle sections. Study of silk formation would help in developing novel fibre spinning techniques. Furthermore, fibroin aggregation is a potential model for amyloidogenesis.

Min Liu¹, Fangfang Yu², Guang Wang³, Tianjing Zhang⁴, and Jing An^4
1the department of Radiology, Beijing Chaoyang Hospital of Capital Medical University, Bei Jing, Bei Jing, China, ²the department of Radiology, Beijing Chaoyang Hospital of Capital Medical University, Bei Jing, China, ³the department of Endocrinology, Beijing Chaoyang Hospital of Capital Medical University, Bei Jing, Bei Jing, China, ⁴MR Collaborations NE Asia, Siemens Healthcare, Bei Jing, Bei Jing, China

The aim of this study was to investigate thyroid T1 value in patients with hypothyroidism (HT). 21 untreated females with clinical hypothyroidism (HT) and 21 healthy control females underwent the thyroid T1 maps with Modified Look-Locker Inversion Recovery technique at 3T MR (Siemens TimTrio). All patients received levo¬thyroxine therapy and were followed after 6 months. The results showed thyroid T1 value at 3T was technically feasible and pretreatment thyroid T1 value increase in HT patients£¬compared with normal woman. After treatment, thyroid T1 value decreased. These suggest T1 mapping may improve disease detection and quantitatively monitor thyroid disease.

Amy R. McDowell¹, Tobias C. Wood², Natalia Petrova¹, Daniele Carassiti¹, Marc Miquel³, David Thomas⁴, Gareth J. Barker², and Klaus Schmierer^5,6
1Blizard Institute, Queen Marys University of London, London, United Kingdom, ²Neuroimaging, King's College London, London, United Kingdom, ³Clinical Physics, Barts Health NHS Trust, London, United Kingdom, ⁴UCL Institute of Neurology, London, United Kingdom, ⁵Barts and The London School of Medicine & Dentistry, Blizard Institute, London, Greater London, United Kingdom, ⁶Neurology, Barts Health NHS Trust, London, Greater London, United Kingdom

Steady-state methods allow time-efficient quantification of relaxation time constants T1 and T2 via the Variable Flip Angle or DESPOT1/2 method. By combining T1 and T2 information, separate water compartments are identified with short, medium and long T2 values, corresponding to myelin water, intra/extra-cellular water, and free water. To validate quantitative MR compartment maps derived from mcDESPOT we applied this technique to post mortem MS spinal cord samples subsequently processed for histology to explore myelin distribution and cellularity. Good correspondence was detected between MWF and myelination, and FWF and cellularity.

Jolanda M Spijkerman¹, Esben T Petersen^1,2, Peter Luijten¹, Jeroen Hendrikse¹, and Jaco J Zwanenburg^1
1Radiology, University Medical Center Utrecht, Utrecht, Utrecht, Netherlands, ²Radiotherapy, University Medical Center Utrecht, Utrecht, Utrecht, Netherlands

The oxygen sensitivity of T1/T2 of cerebrospinal fluid (CSF) may allow non-invasive assessment of oxygen content in ageing and disease. An MLEV pulse sequence was developed for T1/T2 mapping of CSF at high field strength (7 Tesla). A water phantom was used for characterizing B0 and B1 dependency, and reference T1 and T2 maps were made. The T1/T2 were underestimated with the MLEV sequence, but relatively insensitive to B0/B1 variations for B0 offsets < 50Hz and B1 between 80-110%. In vivo results (n=4) were comparable to literature, and showed lower T1/T2 in the peripheral subarachnoid spaces compared to the ventricles.

Nathalie Just¹, Luc Driancourt¹, and Rolf Gruetter^1,2
1CIBM-AIT, EPFL, Lausanne, Switzerland, ²Department of Radiology, Universities of Lausanne and Geneva, Lausanne and Geneva, NA, Switzerland

With the wealth of transgenic mice models allowing preclinical studies of various neurological and oncological disorders, the development of MR quantitative T1 mapping techniques at high field strength is needed. The evaluation of T1 values in various highly resolved anatomical areas of the normal mouse brain with and without manganese using MP2RAGE at 14.1T is proposed.

L. Soustelle^1,2, O. Commowick¹, E. Bannier^1,3, and C. Barillot^1
1Unité VISAGES U746 INSERM-INRIA, IRISA UMR CNRS 6074, University of Rennes, Rennes, France, ²Université de Strasbourg, CNRS, ICube, FMTS, Strasbourg, France, ³Radiology Dept., University Hospital of Rennes, F-35043 Rennes, France

Myelin Water Fraction is a validated measure of myelin density. The proposed approach makes use of recent spatial regularization methods for MWF estimation from clinically compatible acquisitions (typically 11 echoes acquired within 6 minutes). The algorithms were evaluated on both synthetic and clinical data for cross-validation strategy.

Md Nasir Uddin¹, Yushan Huang², Nikolai V Malykhin^1,2, and Alan H Wilman^1
1Biomedical Engineering, University of Alberta, Edmonton, Alberta, Canada, ²Centre for Neuroscience, University of Alberta, Edmonton, Alberta, Canada

Hippocampal subfields are affected selectively in different disease processes but the interpretations are mostly limited to accurate volume measurements. Quantitative T₂ or T₂* relaxometry using high field MRI may be a more sensitive measure for hippocampal subfields compared to volumetric measurements. Here we investigate the absolute measures of spin-echo T₂ and gradient-echo T₂* of the hippocampal subfields at 4.7 T using stimulated echo compensation and susceptibility compensation.

Rafal Panek¹, Liam Welsh¹, Maria A. Schmidt¹, Alex Dunlop¹, Kate L. Newbold¹, Kee Wong¹, Angela M. Riddell¹, Dow-Mu Koh¹, Dualta Mcquaid¹, Shreerang A. Bhide¹, Kevin J. Harrington², Christopher M. Nutting², Georgina Hopkinson³, Cheryl Richardson³, Simon P. Robinson⁴, and Martin O. Leach^1
1Royal Marsden NHS FT and Institute of Cancer Research, Sutton, Surrey, United Kingdom, ²Royal Marsden NHS FT and Institute of Cancer Research, London, United Kingdom, ³Royal Marsden NHS FT, London, United Kingdom, ⁴Institute of Cancer Research, Sutton, Surrey, United Kingdom

In this study, the reproducibility of T2* was calculated, and blood oxygen saturation dependence of tissue relaxation times simulated, for patients with squamous cell carcinoma of the head and neck at 3T. The simulation of tissue T2* shows that the sensitivity of the method increases as a function of blood oxygenation, and is sufficient for the majority of SCCHN oxygen tensions, with the exception of anoxic and severely hypoxic tumour regions where sensitivity is lower than reproducibility coefficient (11%). Both baseline T2* values and blood volumes should be taken into account when employing changes in T2* for assessing tissue oxygenation.

Charles Robert Castets¹, William Lefrançois¹, Aurélien Julien Trotier¹, Emeline Julie Ribot¹, Jean-Michel Franconi¹, and Sylvain Miraux^1
1RMSB - UMR5536, CNRS - Université de Bordeaux, Bordeaux, Aquitaine, France

In this study, a dual cine and 3D T1 mapping method using stack-of-spirals sampling scheme and a Look-Locker approach at high magnetic field is presented. This method has been tested on mice beating heart and allows to differentiate with a high contrast to noise ratio the blood and the myocardium. The dual quantification of T1 and cardiac cinematic might allows a precise characterization of tissue damages induced by a myocardial infarction.