Robert James Anderson¹, Benedikt A. Poser¹, William A. Grissom², and Victor Andrew Stenger^1
1Dept. of Medicine, University of Hawaii, Honolulu, HI, United States, ²Dept. of Biomedical Engineering, Vanderbilt Universiy, Nashville, TN, United States

There is recent interest in Simultaneous Multi-Slice (SMS) imaging because of reduced imaging times. SMS excitation is typically achieved with a single-slice pulse modulated to excite N identical slices. However, there has been little work done on extending SMS excitations to multi-dimensional RF pulses. For example, the “Fast-k_z” or “spokes” 3D RF pulse has been shown to excite thin slices with in-plane B1+ inhomogeneity reduction. We present a simple, analytical “proof-of-concept” SMS Fast-k_z pulse for correcting the central brightening associated with B1+ inhomogeneity from a volume transmitter. We demonstrate the excitation of multiple B1+ inhomogeneity compensated brain slices at 3T.

Jean-Nicolas Dumez¹ and Lucio Frydman^1
1Department of Chemical Physics, Weizmann Institute of Science, Rehovot, Israel

The concepts of spatiotemporal encoding are exploited to design a new class of multidimensional pulses. In particular, 2D pulses are implemented, which operate in a hybrid direct and reciprocal excitation space. These hybrid 2D pulses are shown to be compatible with SPEN as well as Fourier imaging. In the case of Fourier imaging, a self-unfolding mechanism can be used to retrieve the region of interest even when it overlaps with excitation sidebands, thus allowing for larger bandwidths. Phantom experiments at 7T are used to illustrate the properties of SPEN-based multidimensional pulses.

Adam B. Kerr¹, Peder E.Z. Larson², Daniel B. Vigneron², and John M. Pauly^1
1Electrical Engineering, Stanford University, Stanford, CA, United States, ²Radiology, UCSF, San Francisco, CA, United States

A novel approach for designing multiband spectral-spatial RF pulses with complex spectral profile specifications is described. Example RF pulses for application to C-13 imaging are presented and experimentally validated. The complex profile capability is also exploited to achieve a 20% reduction in the peak B1 required by a complex-phase spectral-spatial pulse compared to its linear-phase analog by optimizing the phase for each spectral band.

William A. Grissom¹ and Mark D. Does^1
1Biomedical Engineering, Vanderbilt University, Nashville, Tennessee, United States

A new class of B1+-selective RF pulses is presented that produce excitation only over a prescribed |B1+|, and are designed using the Shinnar-Le Roux algorithm. The pulses' amplitude waveforms resemble refocused gradient trapezoids, while their frequency waveforms resemble sine-modulated slice-selective pulses. Simulated and experimental validation results are presented, including a simulated comparison to an adiabatic (BIR-4) pulse. The pulses could be used for reduced-FOV imaging based on B1+, or in place of non-selective adiabatic pulses, where they would have the advantage of providing a uniform excitation over only a prescribed B1+ range, with a minimum pulse duration.

Robert James Anderson¹, Benedikt A. Poser¹, and Victor Andrew Stenger^1
1Dept. of Medicine, University of Hawaii, Honolulu, HI, United States

Two methods for achieving simultaneous multi-slice (SMS) excitation, multi-band and PINS RF pulses, are combined with a fat-suppressing spectral-spatial pulse in order to minimize the time needed to acquire a single volume of the brain. Both pulses are shown to excite only the water signal across nine simultaneously excited slices of a water/oil phantom and a human brain. Without the need to run a fat-saturating pre-pulse and the almost nine-fold reduction in acquisition times due to SMS, it was possible to acquire a volume with voxel size 3.5x3.5x5mm³ within a TR of 275ms.

Yeji Han¹, Yoojin Lee¹, Dong Chan Kim¹, Joonsoo Kim¹, and HyunWook Park^1
1Department of Electrical Engineering, Korea Advanced Institute of Science and Technology, Daejeon, Daejeon, Korea

Continuously moving table (CMT) MRI is an efficient whole-body (wb) imaging technique for screening-type applications, which provides spatially and temporally continuous images. However, CMT-MRI has to pay careful attention to fat suppression because the rationale behind the conventional fat suppression approaches, including short tau inversion recovery (STIR) and fat saturation, is not applicable to the CMT cases. In this abstract, we propose an expansion of the GR technique for CMT sequences and compare it with other fat suppression techniques to demonstrate the efficiency of the GR algorithm for fat suppression.

Leah C. Henze Bancroft¹, Diego Hernando², Kevin M. Johnson³, Frederick Kelcz⁴, Roberta Strigel⁴, and Walter F. Block^3,5
1Biomedical Engineering, University of Wisconsin - Madison, Madison, WI, United States, ²Radiology, University of Wisconsin-Madison, Madison, WI, United States, ³Medical Physics, University of Wisconsin - Madison, Madison, WI, United States, ⁴Radiology, University of Wisconsin - Madison, Madison, WI, United States, ⁵Biomedical Engineering, University of Wisconsin-Madison, Madison, WI, United States

The IDEAL fat/water decomposition algorithm makes use of multipeak fat model that improves decomposition process with Gradient Echo and Spin Echo sequences. However, when a bSSFP sequence is used the resulting signal magnitude and phase responses can differ significantly from the standard multipeak fat signal model. Presented here are previously uncharacterized changes to this model due to bulk phase changes and amplitude modulations that occur with the use of a bSSFP sequence as well as a method for incorporating this model into a fat/water decomposition algorithm.

Feng Zhao¹, Hao Sun², Jon-Fredrik Nielsen¹, Jeffrey A. Fessler³, and Douglas C. Noll^1
1Biomedical Engineering, The University of Michigan, Ann Arbor, MI, United States, ²EECS, The University of Michigan, Ann Arbor, MI, United States, ³EECS, University of Michigan, Ann Arbor, MI, United States

We propose to use a bSSFP-like imaging sequence, referred to as Small-tip Fast Recovery imaging (STFR), to simultaneously suppress banding artifacts and separate water/fat. This method works for both small and large tip angles, which is advantageous over a recently proposed bSSFP-based fat suppression method called LAMA.

Xiaocheng Wei¹, Yongchuan Lai¹, and Weiwei Zhang^1
1GE Healthcare, Beijing, Beijing, China

The traditional chemical shift selective (CHESS) radiofrequency pulse is often ineffective in suppressing lipid within a large field of view (FOV) due to the presence of B0 inhomogeneity. In this abstract, we propose a new method to suppress lipid signal, Dual-Region Fat Saturation (DRFS). DRFS divides a scan region into multiple sub-regions, in which the shimming could then be applied independently. Volunteer results show that compared with CHESS, DRFS could achieve better and uniform fat suppression across entire FOV, which enable DRFS a much broader application.

Tianjing Zhang^1,2, Ning Jin^1,3, Weiguo Li^1,4, Xiaoming Yin¹, and Andrew C. Larson^1,2
1Radiology Department, Northwestern University, Chicago, Illinois, United States, ²Biomedical Engineering Department, Northwestern University, Evanston, Illinois, United States, ³Siemens Healthcare, Columbus, Ohio, United States, ⁴University of Illinois at Chicago, Chicago, Illinois, United States

This work introduces the combination of free-breathing technology with a fat/water separation method. A revised MGRE sequence was implemented to acquire imaging data during each volunteer’s free-breathing process, followed by an off-line image reconstruction. After that a global optimization algorithm was performed for fat/water separation. The results demonstrated that respiratory self-gated multiple gradient recalled echo sequence with the variable projection fat/water separation method is efficient in reducing the motion artifacts in abdominal imaging. In that way there are no restrictions for slice coverage, spatial resolution as well as scan time in the abdominal fat/water separation applications.

Ken-Pin Hwang^1,2, Olen Rambow², Ersin Bayram³, John D. Hazle², John E. Madewell⁴, Zachary W. Slavens⁵, Anthony T. Vu⁵, and Jingfei Ma^2
1Global Applied Science Laboratory, General Electric Healthcare, Houston, TX, United States, ²Department of Imaging Physics, University of Texas MD Anderson Cancer Center, Houston, TX, United States, ³Global Applied Science Laboratory, GE Healthcare, Houston, TX, United States, ⁴Department of Radiology, University of Texas MD Anderson Cancer Center, Houston, TX, United States, ⁵MR Engineering, General Electric Healthcare, Waukesha, WI, United States

Errors in conventional two-point fat-water separation methods may occur when acquired echo times deviate far from those expected by the separation algorithm. Single-pass, dual-echo sequences are particularly vulnerable when pursuing high resolution at higher field strengths, where the increased frequency shift of lipid demands shorter echo times and echo spacings. This study alleviates this resolution limitation by extending the sampling window to the ramps of the readout pulses and freeing the echo time constraints with a flexible TE algorithm. With this combined method, we demonstrate improved spatial resolution while also improving the overall scan efficiency the sequence.

Jean-Nicolas Dumez¹, Rita Schmidt¹, and Lucio Frydman^1
1Department of Chemical Physics, Weizmann Institute of Science, Rehovot, Israel

A mechanism based on spatiotemporal encoding is exploited to add spectral selectivity to a pair of slice-selective frequency-swept pulses. The SPEN-based SPSP selectivity is used for fat suppression in spin-echo imaging. It does not require fast oscillating gradients and provides high-quality slice profiles. It can also be used to acquire images for two chemical species simultaneously. These concepts are illustrated with phantom experiments at 7T. Examples of fat suppression and water/fat imaging are also shown for breast imaging of human volunteers at 3T.

Axel J. Krafft¹, Brian Allen Taylor¹, Hannah Lin^1,2, Ralf B. Loeffler¹, and Claudia M. Hillenbrand^1
1Radiological Sciences, St. Jude Children's Research Hospital, Memphis, TN, United States, ²Rhodes College, Memphis, TN, United States

Iron overload assessment is one of the most prominent applications of multi-echo GRE-based quantitative T₂^* mapping. One of the major confounding factors arises in the presence of fat due to additional modulations of the mGRE signal. However, these modulations can be modeled and have led to dedicated techniques for fat-water quantification with T₂^* estimation. Here, we systematically analyze a recently proposed autoregressive moving average (ARMA) model for its ability to simultaneously quantify fat-water concentrations and the associated T₂^* times. The ARMA model is compared to conventional fitting approaches and evaluated in phantoms and volunteer data.

Junmin Liu¹, David W. Holdsworth^1,2, and Maria Drangova^1,2
1Imaging Research Laboratories, Robarts Research Institute, Schulich School of Medicine & Dentistry, University of Western Ontario, London, Ontario, Canada,²Department of Medical Biophysics, Shulich School of Medicine & Dentistry, University of Western Ontario, London, Ontario, Canada

A single-point Dixon fat-water separation method is presented to address the concern of relative long scan times associated with multi-echo acquisition. The method derives field maps using the bias-corrected fuzzy C-means (BCFCM) algorithm from a single-echo data set. In vivo 3D experimental results demonstrate that the proposed technique offers robust fat-water separation from a single data set; the results are compared those generated by the three-point IDEAL technique.

Olen Rambow¹, John D. Hazle¹, John Clark², and Jingfei Ma^1
1Imaging Physics Department, University of Texas MD Anderson Cancer Center, Houston, TX, United States, ²Department of Electrical Engineering, Rice University, Houston, TX, United States

The possible solutions to the signal equations in two-point Dixon imaging are the intersection points of two ellipses in the water-fat plane. When the multi-peak spectrum of fat is included in the signal model, the ellipses tilt in such a way that at certain echo time combinations only one solution remains in the region where water and fat are both non-negative. This makes it possible to select the correct solution without phase correction. Simulation, phantom, and in vivo data show that the accuracy achievable by this method depends on both SNR and the echo times.

Chen Cui¹, Xiaodong Wu², Milan Sonka², and Mathews Jacob^2
1Electrical and Computer Engineering, The University of Iowa, Iowa city, IA, United States, ²Electrical and Computer Engineering, The University of Iowa, Iowa city, Iowa, United States

We introduce a novel fat-water decomposition based on non-iterative graph surface estimation.Global optimum of solution is guaranteed through 3D graph searching scheme. Method is applicable to multi-fat peaks situtation. Method are tested on multiple clinical datasets successfully. Computation expense is saved significantly compared to other major algorithms.

Takeshi Yokoo^1,2, Qing Yuan¹, and Ivan E. Dimitrov^2,3
1Radiology, UT Southwestern Medical Center, Dallas, TX, United States, ²Advanced Imaging Research Center, UT Southwestern Medical Center, Dallas, TX, United States,³Phillips Medical Systems, Highland Heights, OH, United States

This proof of concept study shows that a closed form estimation of multipeak proton-density fat fraction is possible using multiecho gradient-echo imaging. It compares favorably to the conventional nonlinear least-square fitting methods.

David S. Smith¹, Johan Berglund², Joel Kullberg³, Håkan Ahlström³, Malcolm J. Avison⁴, and E. Brian Welch^5
1Institute of Imaging Science, Vanderbilt University, Nashville, Tennessee, United States, ²Philips Healthcare, Stockholm, Sweden, ³Department of Radiology, Uppsala University, Uppsala, Sweden, ⁴Department of Pharmacology, Vanderbilt University, Nashville, Tennessee, United States, ⁵Department of Radiology and Radiological Sciences, Vanderbilt University, Nashville, Tennessee, United States

This research describes a robust fat-water separation algorithm for multi-echo MRI applicable to datasets spanning a wide range of anatomy, magnetic field strengths and collected echo times. The algorithm is validated in the context of the 2012 ISMRM Fat-Water Challenge. We show evidence of the power of image-based metrics to predict the best (or nearly best) option among multiple results obtained with various advanced fat-water separation algorithms. No matter the contest outcome, we look forward to sharing the full details of our final algorithm and to seeing many excellent solutions developed by other teams participating in the ISMRM Fat-Water Challenge.

Jingfei Ma¹ and John D. Hazle^1
1Imaging Physics, MD Anderson Cancer Center, Houston, TX, United States

We present a novel tiered, multi-threaded region growing algorithm for two-point Dixon water and fat imaging that can overcome the difficulty in phase correction when regions of large noise, artefacts, or isolated tissues are present. Each thread of processing starts from an initial seed and covers only a high-quality region containing the initial seed. Tiered seed masks serve as a quality metric and allow more trusted regions to be processed first even when completely isolated regions are present. Successful application of the algorithm is demonstrated for processing in vivo two-point Dixon images.

Dinghui Wang¹ and James G. Pipe^1
1Neuroimaging Research, Barrow Neurological Institute, Phoenix, AZ, United States

The recently proposed analytical formula obtains two sets of solutions of the field map, water and fat from images acquired at multiple points with flexible TE increments. We propose a post-processing based on regional iteration and region growing to reliably extract the correct field map from the two solutions. The field map can then be used to recalculate water and fat, and correct imaging artifacts such as blurring in spiral imaging. Results of the implementation with 3D spiral imaging suggest the feasibility of the method for a wide range of field map using short, uneven TE increments.

Rodrigo Moreno¹, Thobias Romu², Olof Dahlqvist Leinhard¹, Magnus Borga², and Ebo de Muinck^1
1Department of Medical and Health Sciences, Linköping University, Linköping, Östergötland, Sweden, ²Department of Biomedical Engineering, Linköping University, Linköping, Östergötland, Sweden

This abstract explores the effect of prefiltering in the estimation of T2* from images acquired through symmetric Dixon imaging. Non-stationary Gaussian noise is removed from 8-point Dixon images acquired from the abdomen. T2* is computed by curve fitting of the in-phase images and the improvement of the estimation is computed through R2. The mean of R2 in is improved with the filtering from 0.73 to 0.75, from 0.84 to 0.89 and from 0.84 to 0.93 for fat- and water-dominant regions and ROIs in the liver respectively. Results suggest that advanced signal model fitting is only necessary in the fat-dominant regions.

Pernilla Peterson¹ and Sven Månsson^1
1Medical Radiation Physics, Malmö, Lund University, Malmö, Sweden

This study aims at investigating the potential gain in precision using large or dual flip angles (FAs) for quantification of fat fraction (FF) in skeletal muscles, compared to a small FA approach. The accuracy and precision within muscle tissue in the lower leg of five lymphedema patients was investigated using small, large, and dual FAs. There was no improvement in precision of FF quantification from using neither a large nor dual FA approach. The large FA approach also resulted in a clear overestimation of the FFs. In conclusion, a small FA approach is preferable for FF quantification in skeletal muscles.

Abraam S. Soliman^1,2, Jing Yuan², Terry M. Peters^1,2, and Charles A. McKenzie^1,3
1Biomedical Engineering, University of Western Ontario, London, Ontario, Canada, ²Imaging Research Laboratories, Robarts Research Institute, London, Ontario, Canada,³Medical Biophysics, University of Western Ontario, London, Ontario, Canada

MRI provides a unique ability to distinguish fat and water signals based on their unique chemical shifts. A successful separation largely relies on the correct estimation of B0 magnetic field inhomogeneities. In this work, we validate a previously proposed approach for field map estimation using datasets provided by the ISMRM challenge for water-fat reconstruction. All the datasets acquired with equally-spaced echo-times were processed by our technique. The average online scoring was 98.78 %, demonstrating that our method can be reliably applied on a wide variety of anatomies in clinical practice.

Paul Begovatz¹, Peter Nowotny¹, Tomas Jelenik¹, Bettina Nowotny¹, Birgit Klüppelholz², Guido Giani², Juergen Bunke³, Michael Roden^1,4, and Jong-Hee Hwang^1
1Institute of Clinical Diabetology, German Diabetes Center, Duesseldorf, Germany, ²Institute for Biometry and Epidemiology, German Diabetes Center, Duesseldorf, Germany, ³Philips Healthcare, Hamburg, Germany, ⁴Department of Metabolic Diseases, University Clinics, Heinrich Heine University, Duesseldorf, Germany

DIXON imaging is increasingly being used for liver fat quantification with a flexible Dual-Echo (mDIXON) technique recently introduced that provides a higher SNR through very short echo times. Tests of mDIXON versus 1H-MRS were conducted in a set of fat phantoms and human subjects. mDIXON showed a good correlation (slope=1.00, r=0.99, p<.001, intercept=4.5%) in the fat phantoms, and between mDIXON and 1H-MRS in vivo hepatic fat (slope=1.07, r=0.96, p<.001, intercept=1.6%). This shows that mDIXON can be used to detect changes in liver fat due to intervention in diabetics, but more work must be done for accurate disease diagnosis.

Andreas Petrovic¹, Clemens Diwoky², Eva Hassler¹, Kathrin Ogris¹, and Eva Scheurer^1
1Ludwig Boltzmann Institute for Clinical-Forensic Imaging, Graz, Austria, ²Institute for Medical Engineering, Graz University of Technology, Graz, Austria

In forensic medicine the morphological characterization of subcutaneous (s.c.) hemorrhage caused by violent events is highly desired. Detailed knowledge of the origin and age of injury can help to reconstruct the sequence of events. However, accurate dating of s.c. hemorrhage is still an unsolved problem. In this study we examined volunteers with artificially created hematomas at several points in time after hematoma creation using the IDEAL fat-water separation technique. We found that with this method excellent delineation of s.c. hemorrhage is possible. Additionally, ROI analysis of water fraction images indicates a systematic decrease of water fraction with time.

Simon Chatelin¹, Marie Humbert-Claude², Philippe Garteiser¹, Valérie Vilgrain^1,3, Bernard E. Van Beers^1,3, Zsolt Lenkei², and Ralph Sinkus^1
1U773-CRB3, INSERM, Université Paris Diderot, Sorbonne Paris Cité, Paris, France, ²Laboratoire de Neurobiologie, ESPCI-CNRS UMR 7637, ESPCI-ParisTech, Paris, France,³Department of Radiology, University Hospitals Paris Nord Val de Seine, Beaujon, Clichy, France

Recent studies suggest a significant influence of type-1 cannabinoïd receptors (CB1R) on the puberty maturation processes. A preliminary study showed a decrease of the hippocampus elasticity after CB1R agonist injection. The aim of this study is to assess the significance and causes of this effect in the hippocampus of juvenile rats. The first step consists in the comparison of hippocampus elasticity and cerebral blood flow (CBF) values, from MR–Elastography (MRE) and Flow-sensitive Alternating Inversion Recovery (FAIR) perfusion imaging respectively, after CB1R activation. In a second step, the sensitivity of MRE to CBF modifications is assessed using the vasodilator nicardipine.

Matthew C. Murphy¹, John Huston¹, Clifford R. Jack¹, Kevin J. Glaser¹, Matthew L. Senjem¹, Jun Chen¹, Armando Manduca², Joel P. Felmlee¹, and Richard Leroy Ehman^1
1Department of Radiology, Mayo Clinic, Rochester, MN, United States, ²Department of Physiology and Biomedical Engineering, Mayo Clinic, Rochester, MN, United States

In this work, we present a novel MR elastography pipeline for measuring regional brain stiffness free of edge-related bias and with high test-retest reliability. Repeatability was measured in 10 healthy volunteers, and the results indicated that typical errors for measuring global and regional brain stiffness were <1% and <2%, respectively. Furthermore, the results show that the mechanical properties of the brain follow a characteristic topography. Considering the lobes of the brain, stiffness is greatest in the occipital lobes, followed by the frontal lobes, the temporal lobes and finally the parietal lobes.

Matthew D.J. McGarry¹, Curtis Jonhson^2,3, Elijah Van Houten⁴, Bradley P. Sutton^3,5, John G. Georgiadis^2,3, John Weaver^1,6, and Keith D. Paulsen^1,7
1Thayer School of Engineering, Dartmouth College, Hanover, NH, United States, ²Department of Mechanical Science and Engineering, University of Illinois at Urbana Champaign, Urbana, Illinois, United States, ³Beckman institute for Advanced Science and Technology, University of Illinois at Urbana Champaign, Urbana, Illinois, United States, ⁴Department of Mechanical Engineering, University de Sherbrooke, Sherbrooke, Quebec, Canada, ⁵Department of Bioengineering, University of Illinois at Urbana Champaign, Urbana, Illinois, United States, ⁶Department of Radiology, Dartmouth-Hitchcock Medical Center, Lebanon, NH, United States, ⁷Norris Cotton Cancer Center, Dartmouth-Hitchcock Medical Center, Lebanon, NH, United States

A method of including spatial information in MR elastography mechanical property reconstruction is presented. Soft prior regularization is implemented through the addition of a penalty term to the objective function of a non-linear inversion algorithm to favor solutions with homogeneous property values in predefined spatial regions. Viscoelastic properties of the gray and white matter of the frontal lobe are presented to demonstrate the technique.

Anthony J. Romano¹, Michael Scheel², Sebastian Hirsch³, Juergen Braun⁴, and Ingolf Sack^2
1Physical Acoustics, Naval Research Laboratory, Washington, DC, United States, ²Department of Radiology, Charite-Universitatsmedizin, Berlin, Germany, ³Department of Radiology, Charité - Universitätsmedizin Berlin, Berlin, Germany, ⁴Institute of Medical Informatics, Charite - University Medicine Berlin, Berlin, Germany

White matter is composed primarily of myelinated axons which form fibrous, organized structures and can act as waveguides for the anisotropic propagation of sound. The evaluation of their elastic properties reguires both knowledge of the orientation of these waveguides in space, as well as knowledge of the waves propagating along and through them. Here, we present waveguide elastography for the evaluation of the elastic properties of the Corona Radiata, In Vivo, using a fusion of MRE, DTI, and anisotropic inversions. It is shown that the Corona Radiata can be characterized by Orthotropic anisotropy.

Curtis L. Johnson^1,2, Matthew D.J. McGarry³, John B. Weaver^3,4, Keith D. Paulsen^3,4, Huan Wang^2,5, William C. Olivero^2,5, Bradley P. Sutton^2,6, and John G. Georgiadis^1,2
1Mechanical Science and Engineering, University of Illinois at Urbana-Champaign, Urbana, IL, United States, ²Beckman Institute for Advanced Science and Technology, University of Illinois at Urbana-Champaign, Urbana, IL, United States, ³Thayer School of Engineering, Dartmouth College, Hanover, NH, United States, ⁴Dartmouth-Hitchcock Medical Center, Lebanon, NH, United States, ⁵Surgery, University of Illinois Medical School, Urbana, IL, United States, ⁶Bioengineering, University of Illinois at Urbana-Champaign, Urbana, IL, United States

Multishot MRE sequences for the brain allow for the acquisition of high-resolution displacement data, which can be used to create shear modulus maps with spatial variations consistent with white matter architecture. Through atlas-based segmentation, the mechanical properties of the corpus callosum and corona radiata are quantified. Healthy subjects are used to show that these structures are distinct from each other and from the total white matter. The reliability of these local measures is demonstrated through multiple scans of a single subject. This work shows that localized differences in brain mechanical properties may be investigated using high-resolution MRE.

Pascal Hagot¹, Emeline Lamain², Tom Doel³, Xavier Maître¹, Redouane Fodil⁴, and Luc Darrasse^1
1IR4M(UMR8081), Univ Paris-Sud, CNRS, Orsay, France, ²Service de Neuroradiologie, Univ Paris-Sud, Hôpital Bicêtre, Le Kremlin-Bicêtre, France, ³Oxford University Computing Laboratory, Univ of Oxford, Oxford, United Kingdom, United Kingdom, ⁴IRMB, Univ Paris XII, INSERM (U955), Creteil, France

Localisation of collapses in the upper airways of patients woth obstructive sleep apnea syndrome is a critical issue to guide and individualize the potential surgical option and procedures. Acoustic measurements may provide valuable information on the upper airway conformation and allow discrimination of healthy and pathological region. Mean amplitudes per slice ranged between 13 and 42 μm along the anterior-posterior direction, 11 and 43 μm along the feet-head direction, 11 and 47 μm along the right-left direction. The generated pressure wave is efficiently guided through the upper airways from the mouth down to the end of the trachea.

Bogdan Dzyubak¹, Armando Manduca², Kevin J. Glaser², Meng Yin², and Richard Leroy Ehman^2
1Mayo Graduate School, Mayo Clinic, Rochester, MN, United States, ²Radiology, Mayo Clinic, Rochester, MN, United States

MR Elastography (MRE) is increasingly being adopted as a practical approach for clinical diagnosis and staging of hepatic fibrosis using noninvasive measurements of liver stiffness. A fully automated algorithm that determines appropriate ROIs from which to report tissue stiffness and agrees with expert readers has been presented previously. Improvements that enable the algorithm to analyze images with motion artifact, signal inhomogeneity, and low edge contrast are presented here. The stiffnesses calculated by the algorithm across 500 cases had a similar discrepancy with clinical readers as the inter-reader discrepancy reported by our previous study.

Ingolf Sack¹, Rolf Reiter¹, Korinna Joehrens², Andreas Fehlner³, Sebastian Hirsch¹, Jing Guo⁴, Rajan Somasundaram⁵, Daniel Seehofer⁶, Carsten Kamphues⁶, Abbas Samani⁷, and Juergen Braun^8
1Department of Radiology, Charité - Universitätsmedizin Berlin, Berlin, Berlin, Germany, ²Department of Pathology, Charité - Universitätsmedizin Berlin, Berlin, Berlin, Germany, ³Department of Radiology, Charité University Medicine, Berlin, Berlin, Germany, ⁴Department of Radiology, Charite - University Medicine Berlin, Berlin, Berlin, Germany, ⁵Interdisciplinary Rescue Center, Charité - Universitätsmedizin Berlin, Berlin, Berlin, Germany, ⁶Department of Transplantation Surgery, Charité - Universitätsmedizin Berlin, Berlin, Berlin, Germany, ⁷Department of Electrical & Computer Engineering, University of Western Ontario, London, Ontario, Canada, ⁸Institute of Medical Informatics, Charite - University Medicine Berlin, Berlin, Berlin, Germany

Elastography has been established for the clinical assessment of liver fibrosis. However, little is known about the relationship between viscoelastic constants and pathophysiological mechanisms in the liver. In this study, samples of the liver of 16 patients with different degrees of fibrosis, inflammation and steatosis were investigated by wide-range MRE and by static indentation for the interaction between mechanical constants and structural parameters of human liver according to histology, matrix protein quantification and function tests. As a main result, MRE was the most sensitive modality to the degree of fibrosis outperforming static indentation experiments and chemical quantification of matrix proteins.

Ramin S. Sahebjavaher¹, Guy Nir¹, Mohammad Honarvar¹, Andrew Yung¹, Piotr Kozlowski¹, Louis O. Gagnon¹, Edward C. Jones¹, Garteiser Philippe², Ralph Sinkus², Silvia Chang¹, Larry Goldenberg¹, and Septimiu E. Salcudean^1
1University of British Columbia, Vancouver, BC, Canada, ²Institut National de la Santé et de la Recherche Médicale (INSERM), Paris, Île-de-France, France

MR elastography (MRE) is performed on ex-vivo prostate cancer specimens. The viscoelastic properties are probed at three frequencies. The reconstructed maps of both the shear (Gd) and loss (Gl) moduli registered to whole-mount histopathology can distinguish cancerous from healthy tissue. The ability to extract frequency dependent information is also assessed.

Jing Guo¹, Sebastian Hirsch², Rolf Reiter¹, Thomas Kroencke¹, Patrick Asbach³, Juergen Braun⁴, and Ingolf Sack^1
1Department of Radiology, Charite - University Medicine Berlin, Berlin, Berlin, Germany, ²Department of Radiology, Charité - Universitätsmedizin Berlin, Berlin, Berlin, Germany, ³MVZ CBF Radiologie, Charite - University Medicine Berlin, Berlin, Berlin, Germany, ⁴Department of Medical Informatics, Charite - University Medicine Berlin, Berlin, Berlin, Germany

3D multifrequency MRE including nonmagnetic shear wave excitation and least-squares multifrequency inversion is introduced for improving the spatial resolution in abdominal-MRE viscoelasticity maps. The new method is demonstrated in health volunteers and patients with ascites and portal hypertension. All three technical key-points, driver, single-shot wave field acquisition and dual parameter inversion, are new in their application to abdominal MRE and contribute to an improved clinical protocol for the precise and spatially highly resolved mechanical assessment of the liver and the spleen.

Anthony J. Romano¹, Ria Mazumder², Seongjin Choi³, Bradley Dean Clymer⁴, Richard White⁵, and Arunark Kolipaka^5
1Physical Acoustics, Naval Research Laboratory, Washington, DC, United States, ²Electrical and Computer Engineering, Ohio State University, Columbus, OH, United States,³Radiology, The Ohio State University, Columbus, OH, United States, ⁴Electrical and Computer Engineering, The Ohio State University, Columbus, OH, United States,⁵Radiology and Internal Medicine, Ohio State University Wexner Medical Center, Columbus, OH, United States

Previously, we implemented a method called Waveguide Elastography in the analysis of the orthotropic elastic parameters of the corticospinal tracts (CSTs) in the brains of five healthy volunteers. Here, we extend this method in the analysis of an ex-vivo, porcine heart. We used our method to evaluate the anisotropic shear coefficients C₄₄and C₅₅ within the three layers of myocardial fibers (i.e. the epicardial, myocardial, and endocardial). We found that there was significant variation within the shear coefficients indicating that the heart is, at a minimum, orthotropic, with stiffness values ranging from 40-60 kPa.

Ria Mazumder¹, Bradley Dean Clymer¹, Richard White^2,3, and Arunark Kolipaka^2,3
1Department of Electrical and Computer Engineering, The Ohio State University, Columbus, OH, United States, ²Deparment of Radiology, The Ohio State University, Columbus, OH, United States, ³Deptarment of Internal Medicine, Division of Cardiology, The Ohio State University, Columbus, OH, United States

Formalin fixed hearts have been actively used in research to study mechanical properties such as stiffness of left ventricular (LV) myocardium. However, the stiffness of the fresh hearts greatly differs from that of the formalin fixed hearts. Mechanical testing has been the gold standard to determine the stiffness of the myocardium. In this study MR Elastography is used to compare the stiffness of LV myocardium between fresh and formalin fixed hearts within the same animal specimen. We have observed that formalin fixed hearts are significantly stiffer than the fresh hearts.

Aniurdh Damughatla¹, Brian Raterman², Orlando P. Simonetti^2,3, Travis Sharkey-Toppen⁴, Ning Jin⁵, Richard White², and Arunark Kolipaka^2,3
1Biomedical Engineering, The Ohio State University, Columbus, Ohio, United States, ²Radiology, The Ohio State University, Columbus, Ohio, United States, ³Internal Medicine, Division of Cardiovascular Medicine, The Ohio State University, Columbus, Ohio, United States, ⁴Davis Heart & Lung Institute, The Ohio State University, Columbus, Ohio, United States, ⁵Siemens Medical Solutions, The Ohio State University, Columbus, Ohio, United States

In aortic aneurysm or Marfan syndrome the stiffness of the aorta varies spatially, however clinical techniques provide global measure of stiffness. Therefore, there is a need for spatial estimation of stiffness of aorta non-invasively. The purpose of the study is to compare the abdominal aortic stiffness obtained using MRI based pulse wave velocity (PWV) measurements against magnetic resonance elastography (MRE) based stiffness measurements in normal volunteers with increase in age. The MRE stiffness values and PWV values plotted against age resulted in similar correlation in the in initial pool of data demonstrating initial feasibility in the same imaging plane.

Marion Tardieu¹, Marie Poirier-Quinot¹, Emeline Lamain¹, Ralph Sinkus², Luc Darrasse¹, and Xavier Maître^1
1IR4M (UMR8081), CNRS, Univ Paris-Sud, Orsay, France, ²Centre de Recherches Biomédicales Bichat-Beaujon (UMR773), CRB3, Inserm, Paris, France

MR-elastography (MRE) aims at characterizing the properties of tissues by motion sensitized phase measurements, and had to cope with the intrinsic motions of the patient itself. A method taking account a correction for both spatial and displacement field component is proposed and is evaluated on a numerical brain phantom: a degradation of a in vivo 3D MRE acquisition (without motion) by 3D rotations between the three spatial encoded directions. The standard deviation of the difference between the reference acquisition phase and the motion simulated one decreased of about a factor 3 to 6 for motion angle of 0.5° to 5°.

Sebastian Papazoglou^1,2, Heiko Tzschaetzsch¹, Juergen Braun³, and Ingolf Sack^1
1Radiology, Charité University Medicine, Berlin, Berlin, Germany, ²NeuroCure Clinical Research Center, Charité University Medicine, Berlin, Berlin, Germany, ³Medical Informatics, Charite - University Medicine Berlin, Berlin, Berlin, Germany

Magnetic resonance elastography (MRE) provides diagnostically significant average viscoelastic parameters in organs with good signal to noise ratio (SNR) such as, e.g. the human liver. Usually, parameter reconstruction in MRE requires the computation of noise amplifying derivatives and also involves the inversion of noisy wave data. Therefore this direct approach is ineffective in case of very low SNR, e.g. in MRE on the human lung. In this study we present an approach based on the wave equation featuring the inverse Laplace operator, which requires no inversion to robustly recover global viscoelastic parameters even in presence of strong noise.

Mikio Suga^1,2, Takayuki Obata², Hajime Ikeda¹, Atsuhisa Koyama¹, Tetsuya Wakayama³, and Riwa Kishimoto^2
1Graduate School of Engineering, Chiba University, Chiba, Chiba, Japan, ²National Institute of Radiological Sciences, Chiba, Chiba, Japan, ³GE Healthcare Japan, Hino, Tokyo, Japan

Magnetic resonance elastography (MRE) is a noninvasive technique for measuring tissue viscoelasticity. Switching-gradient(s)-induced vibration (SGIV) can be used as a mechanical driving mechanism for MRE. The advantage of this approach is that it can be easily adapted for clinical application, but the reproducibility of measurements has not been confirmed. To evaluate reproducibility of MRE scanning with SGIV (MREwSGIV), this study compared shear wave amplitudes in gel phantoms of different weight and measured the viscoelasticity of the human brain at specific mechanical resonance frequencies twice. The results suggest that MREwSGIV enables reproducible measurements of brain elasticity.

Tomokazu Numano^1,2, Junichi Hata³, Kazuyuki Mizuhara⁴, Kouichi Takamoto⁵, Toshikatsu Washio², Hisao Nishijo⁶, Kazuo Yagi¹, and Kazuhiro Homma^2
1Radiological Science, Tokyo Metropolitan University, Arakawa-ku, Tokyo, Japan, ²Institute for Human Science and Biomedical Engineering, National Institute of Advanced Industrial Science and Technology (AIST), Tsukuba, Ibaraki, Japan, ³The University of Tokyo Hospital, Bunkyo-ku, Tokyo, Japan, ⁴Mechanical Engineering, Tokyo Denki University, Adachi-ku, Tokyo, Japan, ⁵Department of Judo Neurophysiotherapy, University of Toyama, Toyama, Toyama, Japan, ⁶System Emotional Science (Physiology), University of Toyama, Toyama, Toyama, Japan

In this work we report on the development of a new technique for non-MSG MR Elastography sequence based upon a gradient-echo type multiecho MR sequence. In contrast to multiecho imaging major benefit of shortening acquisition time, we use the other benefit; separate images are produced from each echo of the train with different TE. When the period of actuator-generated vibration and the multiecho readout gradient lobes are in agreement, the MSG-like effect serves as the maximum. Moreover, the later generated echo is the greater the MSG-like effect (1st<2nd<3rd...), the non-MSG MRE multiecho sequences include images with different MSG-like effects are used to obtain various simultaneously without increasing the acquisition time.

Guangqiang Geng¹, Michael Green¹, Caroline Rae^1,2, Ralph Sinkus³, Roland G. Henry⁴, and Lynne E. Bilston^1,5
1Neuroscience Research Australia, Sydney, NSW, Australia, ²UNSW, Sydney, NSW, Australia, ³Centre de Recherches Biomédicales Bichat-Beaujon, Paris, France,⁴Departments of Radiology and Biomedical Imaging, Neurology, and Bioengineering Graduate Group, University of California, San Francisco, California, United States, ⁵Prince of Wales Clinical School, UNSW, Sydney, NSW, Australia

Magnetic Resonance Elastography (MRE) measures the mechanical properties of the brain in vivo. Most brain MRE studies have assumed isotropy regardless to the anisotropy shown by diffusion tensor imaging (DTI). We combine MRE and DTI to investigate the anisotropic viscoelasticity of human brains. In WM, shear modulus (µ┴=2.44±0.05kPa) and its anisotropy (µFA=0.32±0.01) were significantly greater (ANOVA, p=0.0107 and 0.0011) than in GM (µ┴= 2.02±0.07kPa, µFA=0.24±0.00), consistent with rheological measurements. Also, μFA increased significantly (p<0.0001) from 0.27 to 0.35 when DTI FA increased from 0.26 to 0.54. DTI enhanced MRE enabled a reliable mapping for the anisotropic viscoelasticity of human brains.

Juergen Braun¹, Ralf Lützkendorf², Jing Guo³, Sebastian Hirsch⁴, Andreas Fehlner⁵, Ingolf Sack⁵, and Johannes Bernarding^6
1Department of Medical Informatics, Charite - University Medicine Berlin, Berlin, Germany, ²Department for Biometry and Medical Informatics, Otto-von-Guericke University Magdeburg, Magdeburg, Germany, ³Department of Radiology, Charite - University Medicine Berlin, Berlin, Germany, ⁴Department of Radiology, Charité - Universitätsmedizin Berlin, Berlin, Germany, ⁵Department of Radiology, Charité University Medicine, Berlin, Germany, ⁶Department for Biometry and Medical Informatics, Otto von Guericke University, Magdeburg, Germany

3D multifrequency MRE at 7 T MRI combined with least-squares multifrequency inversion was applied to five healthy volunteers in order to demonstrate the capability of cerebral MRE for providing maps of viscoelastic parameters with high spatial resolution. Our results show that the spatial resolution of viscoelastic parameter maps can approach the resolution of anatomical MR images. Viscoelastic parameters are significantly higher for white matter compared to gray matter. Ultrahigh field 3D multifrequency MRE may be used in the clinic for the mechanical characterization of tumors or neurodegenerative processes.

Sebastian Hirsch¹, Frauke Beyer¹, Sebastian Papazoglou¹, Jing Guo², Juergen Braun³, and Ingolf Sack^1
1Department of Radiology, Charité - Universitätsmedizin Berlin, Berlin, Berlin, Germany, ²Department of Radiology, Charite - University Medicine Berlin, Berlin, Berlin, Germany, ³Institute of Medical Informatics, Charite - University Medicine Berlin, Berlin, Berlin, Germany

Compressible phantoms comprising agarose gel with gas-filled cavities were examined by a rheometer device and MR Elastography (MRE). An effective medium model was used to explain the bulk modulus revealed by the rheometer. MRE was able to detect the relative order of compressibility in phantoms, while severely underestimating the absolute numbers of the bulk moduli. Numerical simulations revealed that compression MRE is sensitive to noise imposing the need of appropriate post-processing. In conclusion, MRE can measure compression properties of tissue, while a more noise-robust inversion method is required for quantitative measurements of bulk moduli.

Loribeth Q. Evertz¹, Jun Chen¹, Jennifer L. Kugel¹, and Richard Leroy Ehman^1
1Mayo Clinic, Rochester, MN, United States

Commercial phantoms for system calibration and validation are common for imaging modalities like CT and MRI, including resolution and flow phantoms. However, there currently are no commercial sources for a suitable liver-mimicking test phantom for MRE. The purpose of this study was to identify and test readily available and inexpensive materials that might be useful as simple test phantoms for hepatic MRE.

Curtis L. Johnson^1,2, Joseph L. Holtrop^2,3, Matthew D.J. McGarry⁴, John B. Weaver^4,5, Keith D. Paulsen^4,5, Bradley P. Sutton^2,3, and John G. Georgiadis^1,2
1Mechanical Science and Engineering, University of Illinois at Urbana-Champaign, Urbana, IL, United States, ²Beckman Institute for Advanced Science and Technology, University of Illinois at Urbana-Champaign, Urbana, IL, United States, ³Bioengineering, University of Illinois at Urbana-Champaign, Urbana, IL, United States, ⁴Thayer School of Engineering, Dartmouth College, Hanover, NH, United States, ⁵Dartmouth-Hitchcock Medical Center, Lebanon, NH, United States

Most estimates of the mechanical properties of brain tissue from MRE are average values computed over only a small number of slices. This is derived from the need to keep scan times short for subject comfort and safety, which leads to acquisitions exhibiting low spatial resolution or poor brain coverage. In this work we take advantage of the SNR efficiency of 3D multislab acquisitions to develop an MRE sequence capable of acquiring high-resolution MRE displacement data with whole-brain coverage in a reasonable scan time.

Phillip J. Rossman¹ and Jun Chen^1
1Mayo Clinic, Rochester, MN, United States

A new method for measuring the complex modulus of soft "tissue-like" materials using the basic principles of Dynamic Mechanical Analysis (DMA) is presented. Results are compared to to those obtiained using Magnetic Resonance Elastography (MRE).

Dieter Klatt¹, Temel Kaya Yasar², Thomas J. Royston¹, and Richard L. Magin^1
1Department of Bioengineering, The University of Illinois at Chicago, Chicago, Illinois, United States, ²Department of Mechanical & Industrial Engineering, The University of Illinois at Chicago, Chicago, Illinois, United States

Sampling interval modulation (SLIM)-MRE is introduced as new recipe for the arrangement of the motion encoding gradients (MEG). SLIM-MRE is independent of the type of MRE sequence and enables complete acquisition of 3D displacements of a monofrequency vibration within a single temporally-resolved MRE experiment. The displacement components are sampled using different time intervals and, in doing so, they are encoded as different apparent frequencies in the MRE signal phase. Thus, all three spatial components are stored in the same k-space and can be decomposed from eight temporally-resolved MRE experiments instead of 12-24 experiments, as is commonly performed in conventional MRE.

John C. Bosshard¹ and Steven M. Wright^1
1Electrical & Computer Engineering, Texas A&M University, College Station, TX, United States

Highly parallel imaging with a 64 channel receive array in a "sandwich" configuration was used to capture MR elastography images of an emerging mechanical wave at 156 x 125 m resolution. The same system was also used to perform simultaneous single echo acquisition imaging at the top and bottom boundaries of the sample using a non-linear gradient coil for compensation of RF coil phase. This method may have applications in microscopic MRE or for monitoring of elastic properties during single-shot events.

Roger C. Grimm¹, Jun Chen¹, Scott A. Kruse¹, and Richard Leroy Ehman^1
1Mayo Clinic, Rochester, MN, United States

Elastography acquisitions measure a propagating complex shear wave field at multiple phase points to estimate stiffness. In general there are 3 polarizations, x, y, z. Using the TR to advance the phase in a multi-frequency acquisition has been previously described. The purpose of this work is to examine timing options where the TR provides the phase advance in interleaved 3 polarization acquisitions. This eliminates misregistration of the wave fields images. When there is no gradient dead time, this results in discrete TE/TR and motion sensitivity values.

Simon Auguste Lambert¹, Simon Chatelin¹, Peter Nasholm², Lauriane Juge¹, Philippe Garteiser¹, Leon Ter Beek³, Valérie Vilgrain¹, Bernard E. Van Beers¹, Lynne E. Bilston⁴, Bojan Guzina⁵, Sverre Holm², and Ralph Sinkus^1
1CRB3- INSERM U773, Univ Paris Diderot, Sorbonne Paris Cité, Clichy, Paris, France, ²Informatics Department, University of Oslo, Oslo, oslo, Norway, ³Philips Healthcare, Best, best, Netherlands, ⁴Neuroscience Research Australia, University of New South Wales, Randwick, NSW, Australia, ⁵Civil Engineering, University of Minnesota, Minneapolis, Mn, United States

Recently, it has been hypothesized that propagation of waves at a macroscopic scale might be influenced by the presence of micro-obstacles and hence lead to an apparent viscosity that can be revealed when exploring the tissue at different frequency. We demonstrated on calibrated phantoms by simulation, experiment, and theory that the frequency-dependence of mechanical shear wave scattering can reveal the underlying micro-architecture even within one single voxel. This technique opens perspective of measuring at the macroscopic level information about micro vasculature of tumors, which is crucial for efficacy monitoring during cancer therapy.

Jihyun Annie Park¹, Andrew Yung², Stefan A. Reinsberg¹, and Piotr Kozlowski^2
1Physics and Astronomy, University of British Columbia, Vancouver, BC, Canada, ²UBC MRI Research Center, Vancouver, BC, Canada

Tumour oxygenation can be measured from changes in lipid T1. Here, we combined the extended two point Dixon and Look Locker technique for a rapid T1 measurement of fat. After simulation studies, the method was first applied in a phantom and in vivo. The T1 recovery curves of fat and water for three regions (subcutaneous fat, external oblique muscle, and femoral muscle) were compared. The T1 values of fat were 442ms and 454ms in subcutaneous fat and external oblique muscle, respectively. We have successfully employed the technique in-vivo to spatially map T1.

Valentina Taviani¹, Kevin M. Johnson¹, Diego Hernando², and Scott B. Reeder^1,3
1Radiology, University of Wisconsin, Madison, WI, United States, ²Radiology, University of Wisconsin-Madison, Madison, WI, United States, ³Medical Physics, University of Wisconsin, Madison, WI, United States

In the presence of hepatic iron overload, quantification of R₂* and fat fraction (FF) can be limited by the rapidly decaying signal. In this work, we propose a multi-echo 3D radial acquisition with angular undersampling and non-Cartesian parallel imaging reconstruction for hepatic R₂* and FF quantification in a single breath hold. Phantom experiments and a preliminary evaluation in healthy volunteers and patients with hepatic iron overload are reported, including a comparison with a hybrid (in-plane radial, through-plane Cartesian) 3D sampling scheme and conventional 3D Cartesian sampling.

Hendrikus Joseph Alphons Crooijmans¹ and Oliver Bieri^1
1Radiological Physics, University of Basel Hospital, Basel, Switzerland

It is generally well accepted that tissue exists of multiple relaxation components, such as for example the free water protons and myelin-bound water protons in brain tissue. However, the spin-echo based quantification of a single T₂ on such tissues is commonly considered being the gold standard. Within such an approach, the echo-spacing is believed to be of influence on the obtained T₂. By means of simulations, we have shown that the echo spacing in a spin-echo acquisition for T₂-quantification does not matter when performing a mono-exponential analysis of a two-pool system.

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

Achieving bone and soft-tissue contrast simultaneously is challenging. The traditional solution consists of co-registering a MRI dataset with a CT dataset; the MRI provides the soft-tissue contrast with high SNR, while CT yields the bone structural information. In this work, a combination of two similar MRI datasets is proposed: quantitative FLASH UTE and Standard FLASH. Bone T1, T2* and proton-density are segmented from the UTE datasets and combined with the standard datasets to form complete quantitative maps. The technique has the advantage of providing the same quantitative information in all tissue types, with three different contrast mechanisms.

Maximilian Freiermuth¹, Linus Willerding², Maximilian F. Reiser¹, Michael Peller¹, and Olaf Dietrich^1
1Josef Lissner Laboratory for Biomedical Imaging, Institute for Clinical Radiology, Ludwig-Maximilians-University Hospital Munich, Munich, Germany, ²Department of Internal Medicine III, Ludwig-Maximilians-University Hospital Munich, Munich, Germany

The purpose of the present study was to optimize and evaluate a fast three-dimensional T1 mapping approach based on the variable flip angle (VFA) technique. The proposed method requires only a single 3D FLASH data set for each time point combined with a longer baseline measurement. The optimal flip angle for the dynamic series can be shown to be α_dyn,opt = arccos[(2E1–1)/(2–E1)], which was also confirmed experimentally. The proposed single-flip-angle technique provides accurate T1 values (within ±5%) in phantom measurements.

Noam Ben-Eliezer¹, Daniel K. Sodickson¹, and Kai Tobias Block^2
1Bernard and Irene Schwartz Center for Biomedical Imaging, Department of Radiology, New York University School of Medicine, New York, NY, United States, ²Bernard and Irene Schwartz Center for Biomedical Imaging, Department of Radiology, NYU Langone Medical Center, New York, NY, United States

T₂ contrast is one of the most clinically useful tools for non-invasive diagnosis and prognosis. Genuine T₂ quantification, however, is impractical due to the long acquisition times associated with full Spin-Echo (SE) acquisitions, or, for fast multi-echo SE sequences, is severely hampered by field inhomogeneities, non-rectangular slice profiles, diffusion effects, and by an inherent bias caused by stimulated echoes. We present a new technique for generating T₂ maps from multi-echo data, based on full Bloch simulation of the experimental pulse-sequence. The technique is assumption free and can further incorporate any experimental factors, e.g., pulse shapes, multiple T₂components and more.

Johannes Tran-Gia¹, Dietbert Hahn¹, and Herbert Köstler^1
1Institute of Radiology, University of Würzburg, Würzburg, Germany

In this work, the previously presented Model-based Acceleration of Parameter mapping (MAP) algorithm for saturation prepared radially acquired datasets is extended for Inversion Recovery (IR) prepared datasets. By incorporating an exponential signal model into the image reconstruction, the proposed IR-MAP algorithm allows quantifying the longitudinal relaxation parameter T1 from a dataset acquired after one single magnetization preparation, leading to extremely short acquisition times of about 7 seconds for one slice of the human brain. The functionality of the algorithm is demonstrated in phantom experiments as well as in-vivo.

Bo Zhao^1,2, T Kevin Hitchens^3,4, Anthony G. Christodoulou^1,2, Fan Lam^1,2, Yijen Wu^3,4, Chien Ho^3,4, and Zhi-Pei Liang^1,2
1Department of Eletrical and Computer Engineering, University of Illinois at Urbana-Champaign, Urbana, IL, United States, ²Beckman Institute, University of Illinois at Urbana-Champaign, Urbana, IL, United States, ³Pittsburgh NMR Center for Biomedical Research, Carnegie Mellon University, Pittsburgh, PA, United States, ⁴Department of Biological Sciences, Carnegie Mellon University, Pittsburgh, PA, United States

A new model-based reconstruction method is presented for accelerating 3D ultra-short echo time (UTE) parametric mapping. The proposed method enables accurate parameter mapping with significantly reduced data acquisition times. It should prove useful for 3D UTE relaxometry and various related applications (e.g., labeled cell tracking).

Kelvin J. Layton^1,2, Mark Morelande¹, David Wright³, Peter M. Farrell¹, Bill Moran¹, and Leigh A. Johnston^1,3
1Electrical and Electronic Engineering, The University of Melbourne, Parkville, Victoria, Australia, ²National ICT Australia, Parkville, Victoria, Australia, ³Florey Neuroscience Institutes, Parkville, Victoria, Australia

This work extends an existing algorithm for multicomponent T2 estimation to account for stimulated echoes using the extended phase graph (EPG) algorithm. The resulting Bayesian algorithm produces reliable multicomponent T2 maps in the presence of dramatic flip angle variation, such as those produced by a transceive surface coil. The method is validated using experimental data acquired on a 4.7T small-bore scanner. Multicomponent T2 maps are generated with and without stimulated echo correction, to demonstrate the importance of the correction. The proposed algorithm is also compared to the commonly used non-negative least squares (NNLS) algorithm and differences are highlighted.

Tri Minh Ngo¹, Haiyan Ding², Mehmet Akçakaya³, Elliot R. McVeigh¹, and Daniel A. Herzka^4
1Biomedical Engineering, Johns Hopkins School of Medicine, Baltimore, MD, United States, ²Biomedical Engineering, Tsinghua University, Beijing, China, ³Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, United States, ⁴Biomedical Engineering, Johns Hopkins University, Baltimore, MD, United States

Myocardial 3D T2 mapping is useful for differentiating between infarct, edema and normal tissue but currently requires prohibitively long acquisition times to be clinically useful. We apply Low-dimensional-structure self-learning and thresholding (LOST) to reconstruct an under-sampled T2 mapping dataset and compare with SENSE reconstruction of equivalent acceleration rates. For rates R3, R3.9, the mean T2 error is lower for LOST than SENSE. At rate R3 SENSE reconstructs edge details better than LOST even though SENSE’s mean T2 error is higher. At rate R3.9, SENSE exhibits high noise amplification while LOST exhibits blurring of high frequency details.

Tao Zhang¹, John M. Pauly¹, and Ives R. Levesque^2
1Electrical Engineering, Stanford University, Stanford, California, United States, ²Radiology, Stanford University, Stanford, California, United States

Parameter mapping can provide intrinsic tissue information to detect pathological changes. Previous studies have shown that compressed sensing with a low-rank constraint can be used to accelerate the lengthy scan time required in parameter mapping. In this work, a locally low rank constraint is applied to parameter mapping. As examples, inversion recovery T1 mapping and multiple-echo T2 mapping are studied. Reconstruction with a locally low rank constraint can provide better accuracy and precision than that with a global low rank constraint.

Bo Zhao^1,2, Fan Lam^1,2, Wenmiao Lu², and Zhi-Pei Liang^1,2
1Department of Eletrical and Computer Engineering, University of Illinois at Urbana-Champaign, Urbana, IL, United States, ²Beckman Institute, University of Illinois at Urbana-Champaign, Urbana, IL, United States

A new model-based reconstruction method is presented to directly reconstruct parameter maps from highly undersampled, noisy k-space data. Some representative results of T2 brain mapping are shown to illustrate the performance of the proposed method. It should prove useful for fast MR parameter mapping with sparse sampling.

Yanqiu Feng¹, Peter David Gatehouse^2,3, David N. Firmin^2,3, Dudley J. Pennell^2,3, Wufan Chen¹, and Taigang He^2,3
1School of Biomedical Engineering, Southern Medical University, Guangzhou, China, ²Cardiovascular Biomedical Research Unit, Royal Brompton Hospital, London, United Kingdom, ³National Heart and Lung Institute, Imperial College, London, United Kingdom

MRI relaxometry mapping (pixel-by-pixel) is sensitive to the noise especially when high-resolution or rapid scans are required. Considering decay signals as the objects to be averaged, we aimed at improving the precision of T2* mapping through selectively Averaging the Decay signals with Similar Underlying Relaxation rates (ADSUR) before curve fitting. This novel algorithm was tested on T2* mapping of simulation data and an ex vivo heart with iron overload. The results demonstrate that the T2* relaxometry mapping can be improved by filtering the serial images with the proposed ADSUR algorithm, independent of curve-fitting models used.

Christopher L. Lankford^1,2 and Mark D. Does^1,2
1Biomedical Engineering, Vanderbilt University, Nashville, Tennessee, United States, ²Vanderbilt University Institute of Imaging Science, Nashville, Tennessee, United States

Fast spin-echo images contain artifacts related to relaxation and imperfect refocusing. These artifacts are explained by well-known signal equations, and the sources of these artifacts can be accounted for if fit as a free parameter during image reconstruction. This abstract demonstrates the potential of this approach for artifact-free T2 mapping in comparison maps derived from Fourier-reconstructed images and voxelwise parameter fits.

Chuan Huang^1,2, Abhishek Pandey³, Tomoe Barr⁴, Ali Bilgin^3,4, and Maria I. Altbach^1
1Department of Medical Imaging, University of Arizona, Tucson, AZ, United States, ²Center for Advanced Radiological Sciences, Department of Radiology, Massachusetts General Hospital, Boston, MA, United States, ³Electrical and Computer Engineering, University of Arizona, Tucson, AZ, United States, ⁴Biomedical Engineering, University of Arizona, Tucson, AZ, United States

Early detection and classification of hepatic tumors and chronic liver disease are two important clinical problems. T2 mapping has been used to improve the characterization of pathologies in the liver. One promising sequence for fast T2 mapping is radial FSE. Most of the reconstruction methods for undersampled radFSE data do not take into account the effects of indirect echoes; this leads to T2 estimates that are dependent on the refocusing pulse slice profile and/or B1 inhomogeneities. Recently, we proposed a reconstruction algorithm – CURLIE (CUrve Reconstruction via pca-based Linearization with Indirect Echo compensation) – combines a principal component model-based algorithm with a slice-resolved extended phase graph signal model. In this work, we demonstrate the ability to obtain accurate T2 maps with indirect echo compensated from liver radFSE data acquired in a single breath hold. The algorithm is also evaluated using phantom and in vivo data. It is also shown that this technique is immune to B1 inhomogeneities and B1 mis-calibration.

Kosuke Morita^1,2, Tomoyuki Okuaki³, Masanori Komi¹, Akiko Kuraoka¹, Daisuke Utsunomiya², Mika Kitajima², Masahiro Hashida¹, and Yasuyuki Yamashita^2
1Radiology, Kumamoto University Hospital, Kumamoto, Kumamoto, Japan, ²Diagnostic Radiology, Faculty of Life Sciences, Kumamoto University Hospital, Kumamoto, Kumamoto, Japan, ³Philips Electronics Japan, 13-37, Kohnan 2-chome, tokyo, Japan

The purpose of our study was to acquire precise 3D T1maps by variable flip angle DESPOT1 method with B1 correction for clinical MR imaging. Moreover, the T1 value of outer-slice regions on 3D acquisition using the ordinary slice selective RF pulse with imperfect rectangular shape is not accurate. To resolve this problem, we evaluated the effect of the B1 inhomogeneity and slice-oversampling factor on 3D T1maps using a phantom and volunteers. Our experiments showed that the T1 value was improved by B1 correction using variable flip angle method and T1 value for 3D T1mapping was improved by optimizing slice-oversampling factor.

Michael Wyss¹, Thomas Kirchner¹, Alex Ringenbach², Klaas P. Prüssmann¹, and Anke Henning^1,3
1Institute for Biomedical Engineering, University and ETH Zürich, Zürich, Switzerland, ²Institute for Medical and Analytical Technologies, University of Applied Sciences of Northwestern Switzerland, Muttenz, Switzerland, ³Max Planck Institute for Biological Cybernetics, Tübingen, Germany

Previously published T1 relaxation times for brain tissue at 7T vary greatly in results and in methods. Only two publications have assessed in vivo T2 relaxation times in the human brain at 7T so far. We present the development and validation of T1 and T2 mapping sequences applicable at 7T in the presence of B1 inhomogeneity. For T1 mapping, a Look-Locker sequence with an adiabatic inversion prepulse and a modified fitting routine was implemented. For T2 mapping the vendor pre-implemented mixed imaging sequence was validated. For cross validation purpose T1 and T2 relaxation times were measured at selected anatomical locations by optimized SV-MRS parameter series.

Tobias C. Wood¹, Samuel A. Hurley², Anthony Vernon³, and Steven C.R. Williams^1
1Neuroimaging, King's College London, Institute of Psychiatry, London, United Kingdom, ²Medical Physics, University of Wisconsin-Madison, Madison, Wisconsin, United States, ³Psychosis Studies, King's College London, Institute of Psychiatry, London, United Kingdom

High-resolution, whole rodent brain, quantitative T1&T2 maps were obtained at 7T using DESPOT techniques, including using a multi-component model to recover the Myelin Water Fraction.

Dylan Breitkreutz^1,2, Kelly C. McPhee¹, R. Marc Lebel², and Alan H. Wilman^1,2
1Department of Physics, University of Alberta, Edmonton, Alberta, Canada, ²Department of Biomedical Engineering, University of Alberta, Edmonton, Alberta, Canada

Stimulated echo compensation enables accurate single component T2 quantification by accounting for both stimulated-echo and spin-echo pathways to produce both T2 and flip angle maps from multi-echo data. In order to avoid non-unique solutions, the algorithm constrains refocusing angles to ≤ 180˚. This work demonstrates the limitations of this assumption, and examines the value of an independent and accurate flip angle map. The results indicate the independent flip angle map is particularly valuable in 2D experiments where refocusing angles exceed 180˚.

Pei-Hsin Wu¹, Cheng-Wen Ko², Ming-Long Wu³, and Hsiao-Wen Chung^1,4
1Department of Electrical Engineering, National Taiwan University, Taipei, Taiwan, ²Department of Computer Science and Engineering, National Sun Yat-Sen University, Kaohsiung, Taiwan, ³Department of Computer Science and Information Engineering, National Cheng Kung University, Tainan, Taiwan, ⁴Institute of Biomedical Electronics and Bioinformatics, National Taiwan University, Taipei, Taiwan

Previous studies demonstrate that DESS provides quantitative T2 imaging. While using 3D acquisition, however, distinct estimated T2 values along section direction are found, particularly at larger flip angles. In this study, we investigate the influence of the uniformity of flip angle distribution in T2 quantification using 3D DESS. Simulation and experiment results suggest that the reliability of T2 quantification across slices within one slab could be improved with appropriate setting of RF flip angle profile, which provides benefit for clinical applications especially when slices comparison is required.

Ece Ercan¹, Peter Börnert², Maarten J. Versluis¹, Tom Geraedts³, Andrew Webb¹, and Itamar Ronen^1
1C.J. Gorter Center for High Field MRI, Department of Radiology, Leiden University Medical Center, Leiden, Netherlands, ²Philips Research Laboratory, Hamburg, Germany,³Philips Healthcare, Best, Netherlands

Ultrashort echo time (UTE) imaging is recently suggested as a potential method to probe the signal from myelin. In this study, we investigate the challenges in UTE relaxometry of the human brain through phantom experiments and UTE acquisitions from the brain white matter. We suggest an improved long T₂ suppression scheme for such experiments. We report the non-exponential behavior that we observed in the relaxometric data from the phantom with a T₂* value of 30 ms as well as in the data from human brain, which is possibly caused by non-fully suppressed tissue constituents with a T₂* value of 10-30 ms (such as myelin water).

Julia Kruisz¹, Andreas Petrovic², Rudolf Stollberger¹, and Eva Scheurer^2
1Institute for Medical Engineering, University of Technology Graz, Graz, Austria, ²Ludwig Boltzmann Institute for Clinical-Forensic Imaging, Graz, Austria

MRI in forensics is becoming increasingly important for a non-invasive and objective documentation before autopsy. However, the usually lower temperatures alter the tissues’ relaxation times which often leads to poor image contrast. In this study we investigated T₁ and T₂ of various tissues from 4° to 38°C. For most investigated samples a linear relationship describes the temperature dependence of relaxation times. Regarding the contrast of a spin echo sequence we found that for certain tissues it is impossible to obtain the same contrast as in vivo by adjusting the repetition time.

Henry H. Ong^1,2, Hua Li^1,3, and John C . Gore^2,4
1Vanderbilt University Institute of Imaging Science, Nashville, TN, United States, ²Department of Radiology and Radiologic Sciences, Vanderbilt University, Nashville, TN, United States, ³Department of Physics and Astronomy, Vanderbilt University, Nashville, TN, United States, ⁴Vanderbilt University, Nashville, TN, United States

Gd-DTPA is a common MRI contrast agent used in a variety of applications such as tumor imaging. Its T₁ relaxivity (r₁) is a fundamental property important for quantitative analyses. Previous studies did not address the behavior of r₁ at ultra-high B₀ or how r₁ may be modified in the presence of realistic media to different extents at different fields which is expected from the Solomon-Bloembergen-Morgan equations. Here, we report Gd-DTPA r₁ measurements as a function of macromolecular content and different high B₀ (up to 15.2T) and found that the effect of macromolecular content disappears at 15.2T in agreement with theory.

Harald Kramer^1,2, Amanda R. Corcos³, Diego Hernando⁴, John F. Berry³, Mark L. Schiebler¹, and Scott B. Reeder^1
1Department of Radiology, University of Wisconsin - Madison, Madison, Wisconsin, United States, ²Institute for Clinical Radiology, Ludwig-Maximilians-University Munich, Munich, Bavaria, Germany, ³Department of Chemistry, University of Wisconsin - Madison, Madison, Wisconsin, United States, ⁴Department of Radiology, University of Wisconsin-Madison, Madison, Wisconsin, United States

Oxygen is paramagnetic and known to influence the relaxation properties of tissues in MRI. Little is known regarding the dependence of oxygen partial pressure on relaxation times (T1 and T2) in fluids. We performed relaxometry experiments in a phantom with varying oxygen partial pressure ranging from 0-760mmHg at clinical field strength. Using measured values of T1 and T2 at different oxygen tensions permitted calculation of the relaxivities (r1 and r2) of oxygen. Differences in oxygen tension can be visualized and quantified with MRI. The results suggest that the injection of oxygen saturated/desaturated saline could be used as a contrast agent.

Dimo Ivanov^1,2, Andreas Schäfer², Andreas Deistung³, Markus N. Streicher², Stefan Kabisch², Ilona Henseler², Elisabeth Roggenhofer², Thies H. Jochimsen⁴, Ferdinand Schweser⁵, Jürgen R. Reichenbach⁵, Kamil Uludag¹, and Robert Turner^2
1Maastricht Brain Imaging Centre, Faculty of Psychology and Neuroscience, Maastricht University, Maastricht, Netherlands, ²Max Planck Institute for Human Cognitive & Brain Sciences, Leipzig, Germany, ³Jena University Hospital - Friedrich Schiller University Jena, Jena, Germany, ⁴Department of Nuclear Medicine, University of Leipzig, Leipzig, Germany, ⁵Medical Physics Group, Jena University Hospital - Friedrich Schiller University, Jena, Germany

The precise knowledge of the oxygenation dependence of the effective transverse relaxation time (T2*) of blood, within the physiologically relevant range will improve the quantitative understanding of the BOLD effect in gradient echo MRI at 7T along the vascular tree. Previous measurements of this dependence have been only reported in vitro and may be confounded by uncompensated susceptibility gradients between the sample and its surroundings. We present an approach of combining T2* and magnetic susceptibility mapping during different respiratory challenges for reliable determination of the change in the T2* of blood with oxygenation in vivo at 7T.

Stefan Zbýn¹, Sebastian Apprich¹, Vladimir Juras¹, Pavol Szomolanyi¹, Sonja M. Walzer², Xeni Deligianni³, Hannes Traxler⁴, Oliver Bieri³, and Siegfried Trattnig^1
1MR Center of Excellence, Department of Radiology, Medical University of Vienna, Vienna, Austria, ²Department of Orthopaedic Surgery, Medical University of Vienna, Vienna, Austria, ³Division of Radiological Physics, Department of Radiology, University of Basel Hospital, Basel, Switzerland, ⁴Center for Anatomy and Cell Biology, Department of Applied Anatomy, Medical University of Vienna, Vienna, Austria

Previous studies on 23Na relaxation times in the intervertebral discs (IVD) reported different values. In this study, we employed novel gradient-echo sequence with variable echo time scheme at 7T for the ex vivo measurements of T1, T2*, fast (T2*F) and slow (T2*S) components of biexponential transversal relaxation times in human lumbar IVDs with different degree of degeneration. Our results suggesting shorter relaxation times in IVDs with higher degree of degeneration compared to less degenerated IVDs. Presented findings may provide the basis for quantification of 23Na content in human IVDs and could help to understand processes associated with degeneration of IVDs.

Mitchell Anthony Cooper^1,2, Thanh Nguyen², Pascal Spincemaille², Martin R. Prince², Jonathan W. Weinsaft³, and Yi Wang^1,2
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

Here we systematically evaluate the inaccuracies of the MOLLI fitting method. We propose a new fitting method, bMOLLI, that utilizes the Bloch equations to model the SSFP signal evolution and account for the flip angle profile and inversion efficiency to provide accurate T1 estimates.

Martin Ott¹, Thomas Benkert¹, Martin Blaimer¹, Peter M. Jakob^1,2, and Felix A. Breuer^1
1MRB Forschungszentrum für Magnet-Resonanz-Bayern e.V., Würzburg, Germany, ²Department of Experimental Physics 5, Institute of Physics, Würzburg, Germany

A comparison of different VFA-methods for T1-mapping is shown. This results that none of the common techniques can produce reliable data.

Sofia Chavez^1,2, Stephen Kish^1,2, Tina McCluskey¹, and Nancy Lobaugh^1,3
1Research Imaging Centre, Centre for Addiction and Mental Health, Toronto, ON, Canada, ²Psychiatry, University of Toronto, Toronto, ON, Canada, ³Neurology, University of Toronto, Toronto, ON, Canada

The Method of Slopes (MoS), which accounts for spatial variations in flip angle, has been shown to yield accurate in vivo 3D T1 maps in brain at 3T. However, several factors may cause variations in the measurement within a subject across time. In this work, we assess the intra-subject reproducibility of T1 mapping with the MoS by repeating scans on five healthy subjects, at three time points on two consecutive days: morning, afternoon, morning. We find very small intra-subject coefficients of variation (CV<8%) in all brain regions, making these T1 maps suitable for the detection of short-term effects(eg. drug-induced).

Mekala R. Raman¹, Yunhong Shu¹, John D. Port¹, Jan-Mendelt Tillema¹, Istvan Pirko¹, Clifford R. Jack¹, and Kejal Kantarci^1
1Mayo Clinic, Rochester, MN, United States

Our objective was to determine the effects of fixation on T1 relaxation of gray matter and white matter in ex vivo brain MRI for optimization of MPRAGE and DIR sequences. We found that T1 relaxation times of fixed brain tissue attenuate over time. Effects of formalin exposure on tissue relaxation time appear to be greater for gray matter than white matter.

Antoine Lutti¹ and Nikolaus Weiskopf^1
1Wellcome Trust Centre for Neuroimaging, Institute of Neurology, University College London, London, United Kingdom

The accuracy of T1 maps obtained with FLASH-based Variable Flip Angle (VFA) methods is affected by residual transverse coherences, spatial inhomogeneities of the transmit field B1 and non-linearities of the RF transmit chain. Here we address all these issues simultaneously and present an acquisition protocol with very high accuracy and precision. The accuracy was improved by ~40% in-vivo compared to a conventional VFA protocol, enabling cutting-edge applications such as mapping of myeloarchitecture.

Mathieu Boudreau¹, Nikola Stikov¹, and Bruce G. Pike^1
1Montreal Neurological Institute, McGill University, Montreal, Quebec, Canada

A recent study reported that three commonly used methods for T₁ mapping (Inversion Recovery, Look-Locker, Variable Flip Angle) measure similar T₁ values in phantoms, but disagree in vivo at 3T (VFA overestimated and LL underestimated relative to IR). This work investigates possible confounding factors that may explain the T₁ trends observed in vivo. Bloch simulations tested the effects of inaccurate B₁ mapping and spoiling on T₁ for all three sequences. These simulations predict a systematic bias in VFA and LL due to these effects, consistent with trends observed in vivo, highlighting the importance of proper calibration with the IR gold standard.

Ryota Sato¹, Toru Shirai¹, Yo Taniguchi¹, Yoshihisa Soutome¹, and Yoshitaka Bito^1
1Central Research Laboratory, Hitachi, Ltd., Kokubunji, Tokyo, Japan

A method for susceptibility-weighted imaging, which can enhance contrast of regions with high susceptibility in arbitrary slice orientation, was developed. To enhance contrast in arbitrary slice orientation clearly, this method uses a susceptibility map estimated by L1 norm regularization. In contrast to conventional methods, the proposed method can enhance contrast of veins and iron depositions without causing streaking artifacts in arbitrary slice orientation. A numerical simulation and experiments on healthy volunteers show this method can enhance contrast of regions with high susceptibility in arbitrary slice orientation.

Mai Murashima¹, Tomohiro Ueno¹, and Naozo Sugimoto^1
1Human Health Science, Graduate School of Medicine, Kyoto University, Kyoto, Japan

In Quantitative susceptibility mapping (QSM), measured magnetic field distributions have to be deconvolved with a dipole field. The dipole field is effective even in a long range, and a rapidly changing function of a position in a short range. In order to access these problems, we performed numerical simulations taking into account partial volume effects in QSM by using the dipole fields with various field sizes and oversampling factors. We found that the diameter 21voxels was enough for the field size, and that the oversampled dipole field of factor 3 was sufficient for observing effects on the susceptibility estimation.

Olaf Dietrich¹, Johannes Levin², Armin Giese³, Annika Plate², Kai Bötzel², Maximilian F. Reiser^1,4, and Birgit Ertl-Wagner^4
1Josef Lissner Laboratory for Biomedical Imaging, Institute for Clinical Radiology, Ludwig-Maximilians-University Hospital Munich, Munich, Germany, ²Department of Neurology, Ludwig-Maximilians-University Hospital Munich, Munich, Germany, ³Center for Neuropathology and Prion Research, Ludwig-Maximilians-University Hospital Munich, Munich, Germany, ⁴Institute for Clinical Radiology, Ludwig-Maximilians-University Hospital Munich, Munich, Germany

The purpose of the present study is to demonstrate the different behaviors of ferric and ferrous iron ions in MRI and to suggest a technique to differentiate quantitatively between both forms of iron in tissue. A phantom consisting of tubes with different concentrations of ferrous and ferric chloride solutions was examined on a 3-Tesla MRI system. A multi-echo gradient-echo was used for both T2* and quantitative susceptibility measurements. Ferrous and ferric chloride show markedly different relaxation behaviors in MRI, but similar influences on the susceptibility. These properties can be used to differentiate ferrous and ferric samples.

Sagar Buch¹, Saifeng Liu¹, Yu-Chung Norman Cheng², and Ewart Mark Haacke^1,2
1McMaster University, Hamilton, Ontario, Canada, ²Academic Radiology, Wayne State University, Detroit, Michigan, United States

Susceptibility mapping is generally used to provide us with the distribution of magnetic susceptibility properties of the different tissues. This abstract introduces a new concept of mapping the susceptibility inside brain sinuses, which are usually discarded from the MR phase. One of our key advances is to include the tissues in the skull outside the sinuses in phase images, before applying the inverse filter. Because the susceptibility effects between air and tissue are so large, we are able to use short effective echo times, produced by complex division of the datasets at two different echo times, to extract the susceptibility.

Issel Anne L. Lim^1,2, Andreia V. Faria¹, Xu Li^1,2, Johnny T.C. Hsu¹, Raag D. Airan¹, Susumu Mori^1,2, and Peter C.M. van Zijl^1,2
1Radiology, Johns Hopkins University, Baltimore, Maryland, United States, ²F.M. Kirby Research Center for Functional Brain Imaging, Kennedy Krieger Institute, Baltimore, Maryland, United States

Quantitative Susceptibility Mapping (QSM) methods have correlated magnetic susceptibility and gray matter (GM) brain iron concentration. Typically, average susceptibility per structure is determined via manual delineation of ROIs, which is labor-intensive and subject to human error. We created a QSM template as part of the Eve atlas, along with an "Everything" Parcellation Map ("EvePM") containing deep GM ROIs drawn from QSM and white matter ROIs from DTI maps. Using automated segmentation, we derived a susceptibility-iron calibration curve for deep gray matter based on known iron values for three regions, which was used to determine iron concentration in six other regions.

Florian Lippus¹, Kerstin Brocker², Jens Groebner¹, Peter Hallscheidt³, and Michael Bock^1
1Dept. of Diagnostic Radiology / Medical Physics, University Medical Centre Freiburg, Freiburg, BW, Germany, ²Dept. of Gynaecology, Heidelberg University Hospital, Heidelberg, BW, Germany, ³Diagnostic and Interventional Radiology, Heidelberg University Hospital, Heidelberg, BW, Germany

Partially absorbable surgical mesh implants can be used to treat Pelvic organ prolapse, which is a common disease in elderly women. When it comes to post-operative complications, MRI could be used to qualify movement or deformation of the implant, but current available mesh implants are not visible in MR images. In this work MR measurements are presented to develop and optimize an MR-visible mesh-implant made of threads with magnetic particles. Particle concentrations were determined based on relaxometric evaluation, and thread material was tested for its MR visibility.

John T. Spear^1,2, Zhongliang Zu^2,3, and John C. Gore^3,4
1Department of Physics and Astronomy, Vanderbilt University, Nashville, TN, United States, ²Vanderbilt University Institute of Imaging Science, Nashville, TN, United States, ³Department of Radiology, Vanderbilt University, Nashville, TN, United States, ⁴Vanderbilt University, Nashville, TN, United States

A new method is reported for quantifying the spatial scales of susceptibility variations based on the effect of diffusion through internal gradients on the spin-lattice relaxation rate in the rotating frame R1rho (=1/T1rho). Model systems with internal field gradients exhibited significant dispersions in R1rho and fastidious analysis of these dispersions revealed three important locking fields. Three images were subsequently taken at these locking fields and combined in a novel image subtraction technique to delineate the average spatial frequency of magnetic inhomogeneities. This technique has the potential to quantitatively characterize magnetically inhomogeneous regions of tissue such as microvasculature in vivo.

Matthew Cronin¹, Samuel J. Wharton¹, and Richard W. Bowtell^2
1Sir Peter Mansfield Magnetic Resonance Centre, The University of Nottingham, Nottingham, Notts, United Kingdom, ²Sir Peter Mansfield Magnetic Resonance Centre, University of Nottingham, Nottingham, Notts, United Kingdom

Although correct mathematical expressions for calculating field perturbations due to anisotropic magnetic susceptibility have been described, its effects have often been modelled using a simplified approach in which the anisotropy is represented by allocating the material an isotropic susceptibility whose magnitude depends on orientation to the field. Here we demonstrate using theory and experiment that this approximation leads to errors in predicting the frequency perturbation due to anisotropic structures. We also show that correct calculations predict interesting behaviour in a hollow cylinder model of the myelin sheath, and go on to demonstrate this behaviour experimentally in a simple model system.

Andreas Petrovic¹, Ferdinand Schweser², Andreas Deistung³, Eva Scheurer¹, and Jürgen R. Reichenbach^2
1Ludwig Boltzmann Institute for Clinical-Forensic Imaging, Graz, Austria, ²Medical Physics Group, Institute of Diagnostic and Interventional Radiology I, Jena University Hospital – Friedrich Schiller University Jena, Jena, Germany, ³Medical Physics Group, Institute of Diagnostic and Interventional Radiology I, Jena University Hospital - Friedrich Schiller University Jena, Jena, Germany

Dating of hemorrhage is important in clinical and forensic medicine. In this study we measured the susceptibility of artificially created hematomas in 6 volunteers at different times to investigate if susceptibility changes due to changes of the oxygenation state of hemoglobin could be used. Quantitative susceptibility mapping is a unique method to measure magnetic susceptibility in vivo. The suspected increase of magnetic susceptibility could not be observed. This could be caused by blood being already fully deoxygenated at the time of the first measurement. Additionally, confounding factors as an increase of the hematocrit and recurring fat during resorption influence the measured susceptibility values.

Se-Hong Oh¹ and Jongho Lee^1
1Department of Radiology, Perelman School of Medicine, University of Pennsylvania, Philadelphia, Pennsylvania, United States

Recent MRI studies have demonstrated that the relative orientation of white matter fibers to the B0 field significantly affects R2* measurement. In this work, the origin of this effect was investigated by measuring R2 and R2* in multiple orientations and fitting the results to a magnetic susceptibility-based model and a magic angle-based model. To further explore the source of magnetic susceptibility effect, the contribution of tissue iron to the orientation dependent R2* contrast was investigated. Additionally, R2* was measured in basal ganglia area to confirm there is no orientation dependence in deep gray matter. Our results suggest that myelin is a primary source for R2* contrast because of its highly oriented structure and large susceptibility value.

Samuel J. Wharton¹ and Richard W. Bowtell^1
1Sir Peter Mansfield Magnetic Resonance Centre, School of Physics and Astronomy, University of Nottingham, Nottingham, United Kingdom

Frequency difference mapping is a recently proposed gradient-echo-based MRI technique for creating image contrast that is sensitive to the orientation of myelinated nerve fibers in white matter. Frequency difference maps (FDM) are created by calculating the difference in phase-based frequency maps acquired at short and long echo times. In this study, we use simulations to determine the optimum TE values for creating in vivo FDM at 7T using two- and three-echo acquisition schemes. Both approaches yield high quality FDM in a 6 minute scan time. The two-echo scheme yields higher SNR, but requires the use of a high-pass filtering step.

Matthew Cronin¹, Samuel J. Wharton¹, Richard W. Bowtell², and Penelope A. Gowland^2
1Sir Peter Mansfield Magnetic Resonance Centre, The University of Nottingham, Nottingham, Notts, United Kingdom, ²Sir Peter Mansfield Magnetic Resonance Centre, University of Nottingham, Nottingham, Notts, United Kingdom

Peripheral rings surrounding white matter (WM) lesions in multiple sclerosis are often analysed using phase images from T2*-weighted, gradient echo scans. The usefulness of this approach is limited by the non-local, dipolar nature of phase contrast. Quantitative susceptibility mapping (QSM) can be used to overcome this problem. Here, by comparison of images and analysis of 1D profiles of signal variation with respect to the distance from the lesion edge, we show that QSM offers a more accurate representation of WM lesions, which reflects the underlying tissue composition more closely than phase or T2*-weighted magnitude data.

Julien Bouvier^1,2, Sebastien Castellani¹, Clément S. Debacker^1,3, Nicolas Pannetier⁴, Irène Troprès^1,5, Alexandre Krainik¹, and Emmanuel Luc Barbier^1
1INSERM U836, Grenoble Institute of Neurosciences, Grenoble, France, ²Philips Healthcare, Suresnes, France, ³Bruker Biospin MRI, Ettlingen, Germany, ⁴University of California San Francisco, San Francisco, CA, United States, ⁵Plate-forme IRMaGe, UJF – INSERM US17 – CNRS UMS 3552, Grenoble, France

Quantitative estimates of the tissue blood oxygen saturation (StO2) may be obtained using a multiparametric quantification of the blood oxygen level dependent effect (multiparametric qBOLD). This method, based on a model of the MR signal, yielded promising experimental results on rodents. However, the first estimates of StO2 obtained in humans with this method matched those reported in the literature for gray matter only. To obtain reliable StO2 estimates in white matter, this study evaluates a solution to account for the bias on StO2 estimate induced by myelin, a paramagnetic substance, using a numerical simulation approach.

Diana Khabipova¹, Yves Wiaux^2,3, Rolf Gruetter^4,5, and José P. Marques^1,5
1LIFMET, EPFL, Lausanne, Switzerland, ²Electrical Engineering, EPFL, Lausanne, Switzerland, ³Radiology, UniGe, Geneva, Switzerland, ⁴LIFMET, École Polytechnique Fédérale de Lausanne, Lausanne, Switzerland, ⁵Radiology, UNIL, Lausanne, Switzerland

Phase imaging has been demonstrated to achieve a good contrast between and within brain tissues at 7T. However, phase imaging suffers from a non-local contrast variation which can be overcome by calculating the underlying magnetic susceptibility maps. As this problem is ill-posed, many regularization methods have been proposed over the past years. In this abstract we do a thorough comparison of some of these methods, focus on the impact of the prior information on the reconstructed susceptibility maps and propose a method to evaluate the quality of the susceptibility reconstruction in the absence of a ground truth.

Johannes Lindemeyer¹, Ana-Maria Oros-Peusquens¹, Kaveh Vahedipour¹, and Nadim Jon Shah^1,2
1Institute of Neuroscience and Medicine 4, Forschungszentrum Jülich, Jülich, Germany, ²Department of Neurology, Faculty of Medicine, JARA, RWTH Aachen University, Aachen, Germany

Unwrapping of MR phase data becomes very time consuming for high-resolution data acquired at ultra-high fields. We present a new algorithm, URSULA (UnwRap of SUbdivided Large Arrays), an approach splitting up the original matrix in smaller-sized 3D volumes which are unwrapped individually. The computed subsets are assembled into a whole volume result using a quality-based approach. Sequential or parallel computing is applicable, especially the latter allowing for a dramatic gain in computing speed. Simulations show good reliability for a moderate number of subsets. The performance of the algorithm is exemplified on a human in vivo dataset measured at 9.4T.

Ferdinand Schweser^1,2, Andreas Deistung¹, Martin Stenzel³, Hans-Joachim Mentzel³, and Jürgen R. Reichenbach^1
1Medical Physics Group, Institute of Diagnostic and Interventional Radiology I, Jena University Hospital - Friedrich Schiller University Jena, Jena, Germany, ²School of Medicine, Jena University Hospital - Friedrich Schiller University Jena, Jena, Germany, ³Section Pediatric Radiology, Jena University Hospital - Friedrich Schiller University Jena, Jena, Germany

Susceptibility weighted imaging (SWI) is a novel imaging technique that exploits the MR phase information to delineate venous vessels and brain lesions. Various approaches have been presented to avoid severe off-resonance artifacts in SWI due to the strong field gradients induced, e.g., by the air-tissue interfaces above the nasal cavities. However, all of these techniques involve rather complex and laborious data processing based on raw complex-valued MR images, which are usually not available in commercial SWI implementations. In this contribution we present a simple and robust post-processing technique that allows eliminating the wrap-artifacts in SWI. The technique may be applied both retrospectively on filtered phase images (commercial SWI) and prospectively on raw complex-valued images.

Saifeng Liu¹, Sagar Buch¹, and Ewart Mark Haacke^1,2
1School of Biomedical Engineering, McMaster University, Hamilton, Ontario, Canada, ²Department of Radiology, Wayne State University, Detroit, Michigan, United States

The quality of Quantitative Susceptibility Mapping (QSM) relies on the accuracy in background field removal. In the SHARP method, phase unwrapping is generally required, which is usually a time-consuming procedure. It could be avoided by using the Laplacian of the field but there are errors in regions with sharp phase changes. Besides, using the Laplacian may lead to reduced accuracy in the processed phase images. We proposed a method which allows for using an optimal kernel size without explicit phase unwrapping. This helps to reduce the processing time and improve the accuracy in background field removal using SHARP.

Huan Tan¹, Ying Wu^1,2, Ryan Hutten¹, Liu Tian³, Yi Wang³, Pottumarthi Vara Prasad^1,2, Issam Awad², and Robert R. Edelman^1,4
1NorthShore University HealthSystem, Evanston, IL, United States, ²The University of Chicago Pritzker School of Medicine, Chicago, IL, United States, ³Weill Cornell Medical College, New York, NY, United States, ⁴Northwestern University Feinberg School of Medicine, Chicago, IL, United States

Susceptibility weighted imaging (SWI) was shown to have a high sensitivity for detecting iron-rich cerebral cavernous malformation lesions. However, SWI is a qualitative technique that does not provide a means to evaluate changes in iron distribution within individual lesions. Quantitative susceptibility mapping (QSM) allows quantitative evaluations of intra-lesional susceptibility changes related to leakage of iron-containing blood products. In this study, we performed QSM in nine patients with CCM to characterize lesion burden, and demonstrate the potential of QSM as a quantitative imaging marker for monitoring disease progression and/or responses to treatments.

Huan Tan¹, Jon Thacker², Tian Liu³, Yi Wang³, and Pottumarthi Vara Prasad^1,4
1NorthShore University HealthSystem, Evanston, IL, United States, ²Northwestern University, Evanston, IL, United States, ³Weill Cornell Medical College, New York, NY, United States, ⁴The University of Chicago Pritzker School of Medicine, Chicago, IL, United States

The magnetic susceptibility of tissue can be measured using quantitative susceptibility mapping (QSM), a novel MRI technique that utilizes a 3D multi-echo gradient echo sequence for data acquisition. Since the quantification of susceptibility is inherently based on the phase information and necessitates 3D acquisition, respiratory motion remains a key challenge. In this study, we implemented a prospective cross-pair navigator method to minimize motion effects. The technique was tested in five volunteers, and the preliminary analysis supports feasibility of acquiring QSM in vivo in human kidneys.

Martin Krämer¹, Andreas Deistung¹, Ferdinand Schweser¹, and Jürgen R. Reichenbach^1
1Medical Physics Group, Institute of Diagnostic and Interventional Radiology I, Jena University Hospital - Friedrich Schiller University Jena, Jena, Germany

The ability to perform phase correction during PROPELLER-EPI image reconstruction is highly advantageous making PROPELLER-EPI a robust segmentation scheme for fast high resolution echo planar imaging. However, only modified phase information is obtained after the complete image reconstruction, hampering further processing of the image phase. To avoid this issue and to make SWI processing of PROPELLER-EPI data possible we propose to perform the processing of phase information on the blade level, before image phase is modified in any way. We show that this reconstruction scheme enables high resolution susceptibility weighted imaging using a fast EPI readout.

Wei Feng¹, Yang Xuan¹, and Ewart Mark Haacke^1
1Radiology, Wayne State University, Detroit, Michigan, United States

Gradient echo sampling of FID and echo (GESFIDE) and its variants have been studied to quantify tissue relaxation rates, including R2, R2* and R2’. In this abstract, we expand this sequence to also achieve susceptibility weighted imaging (SWI), T1W imaging, and MR angiography. Gradient echoes are placed before the refocusing pulse, after the refocusing pulse and even after the spin echo. A carefully selected flip angle, echo time and repetition time allow the unwrapping of multiecho phase images to generate improved SWI images. Furthermore, good T1 contrast can be achieved at short GRE echoes while MRA can be obtained at the spin echo.

Wen-Tung Wang¹, Ningzhi Li², Dzung Pham², and John Anthony Butman^1,2
1Radiology and Imaging Sciences, Clinical Center of the National Institutes of Health, Bethesda, MD, United States, ²Center for Neuroscience and Regenerative Medicine, Bethesda, MD, United States

Susceptibility weighted imaging using segmented EPI dramatically accelerates aquisition as compared with traditional 3D GRE methods. In the context of TBI, the ability to cover the whole brain in 90 seconds with segmented EPI with microhemorrhage detection sensitivity comparable to the 9 minute 3D GRE method makes it attractive for routine clinical use.

Jeam Haroldo Oliveira Barbosa¹, Saifeng Liu², Jin Tang³, Manju Liu⁴, Weili Zheng⁴, Ewart Mark Haacke^2,4, and Carlos Ernersto Garrido Salmon^1
1Department of Phisics, University of Sao Paulo, Ribeirao Preto, Sao Paulo, Brazil, ²School of Biomedical Engineering, McMaster University, Hamilton, Ontario, Canada, ³MRI Institute for Biomedical Research, Detroit, Michigan, United States, ⁴Wayne State University, Detroit, Michigan, United States

There are a variety of means by which iron content in vivo tissue can be measured. These include but are not limited to measuring: transverse relaxation rates (R2 and R2*) and quantitative magnetic susceptibility (QSM). However, the dependency of transverse relaxation rates with water content can confound the accuracy of iron quantification if R2 or R2* are used. To overcome this limitation, magnetic susceptibility mapping has been suggested to evaluate iron deposits in neurodegenerative diseases. In this study, we evaluated the correlation between iron concentration with R2 and R2* and magnetic susceptibility in vivo in the midbrain region for seven healthy elderly subjects.

Jon Thacker¹, Huan Tan², Shivraman Giri³, and Pottumarthi Vara Prasad^2
1Biomedical Engineering, Northwestern, Evanston, Illinois, United States, ²Northshore University Health System, Evanston, Illinois, United States, ³Siemens Healthcare, Chicago, Il, United States

Renal oxygenation is typically assessed through R2* maps. However, it is well known that R2* is sensitive to physiological changes other than oxygenation. R2* is believed to be made up of two components: R2 and R2’. R2’ has been assessed as a potentially more specific indicator of oxygenation. Previous work has been done with asymmetric spin echo (ASE) EPI sequences for acquiring R2’ maps. These have been limited in spatial resolution and confined to breath-hold acquisitions. We have integrated a navigator gated, ASE segmented EPI sequence for increased resolution and removal of the inherent constraints with breath-hold scans.

Karin Markenroth Bloch^1,2, Stefanie Eriksson³, Roger Siemund^4,5, Håkan Sjunnesson^4,5, Emelie Lindgren², and Danielle van Westen^4,5
1Clinical science, Philips, Lund, Sweden, ²Dept. of Medical Radiation Physics, Lund University, Lund, Sweden, ³Dept. of Chemistry, Lund University, Lund, Sweden, ⁴Center for Medical Imaging and Physiology, Skane University Hospital Lund, Lund, Sweden, ⁵Dept. of Diagnostic Radiology, Lund University, Lund, Sweden

The aim of this study was to quantitatively compare two techniques for susceptibility weighted imaging (PADRE and SWI) with regard to hemorrhagic lesion detection and contrast of hemorrhages, as well as to qualitatively compare the image quality.

Ken-Pin Hwang¹, Marcel Warntjes², R. Jason Stafford³, Wolfgang Stefan³, Edward F. Jackson³, John E. Madewell⁴, John D. Hazle³, Zachary W. Slavens⁵, and Tzehping L. Chi^4
1Global Applied Science Laboratory, General Electric Healthcare, Houston, TX, United States, ²Center for Medical Imaging Science and Visualisation, Linköping University, Linköping, Sweden, ³Department of Imaging Physics, University of Texas MD Anderson Cancer Center, Houston, TX, United States, ⁴Department of Radiology, University of Texas MD Anderson Cancer Center, Houston, TX, United States, ⁵MR Engineering, General Electric Healthcare, Waukesha, WI, United States

Brain water content may be assessed with a proton density measurement as derived from a multiple gradient echo sequence, where all causes of signal inhomogeneity are properly corrected. Typically, the effective flip angle of the transmit field and T1 of tissues are mapped with multiple dedicated sequences. Here, we base our corrections on parameters produced from a single multi-parameter mapping technique. We also extrapolate the gradient echo signal using a spectral modeling technique instead of fitting the signal magnitudes. Thus absolute water content assessment is performed with two complementary sequences which provide other quantitative measurements as well.

Olivier M. Girard¹, Virginie Callot², Alexandre Vignaud³, Gopal Varma⁴, Patrick J. Cozzone⁵, David C. Alsop⁴, and Guillaume Duhamel^2
1CRMBM UMR 7339, CNRS / Aix-Marseille Université, Marseille, France, ²CRMBM UMR 7339, Aix-Marseille University, Marseille, France, ³Siemens Healthcare, Saint-Denis, France,⁴Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, MA, United States, ⁵CRMBM UMR 7339, Aix-Marseille Université, Marseille, France

Myelin specific imaging has become an abounding research area because of the high clinical relevance of myelin-associated diseases such as multiple sclerosis (MS). Recently a novel MT contrast, known as inhomogeneous MT - ihMT, has demonstrated high specificity toward brain WM tissue. In this work an ihMT sequence was developed on a 1.5T clinical scanner and applied on the human cervical spinal cord. This new endogenous contrast demonstrated tremendous specificity towards spinal cord WM. This holds great promise for future clinical applications on SC pathologies such as trauma or MS.

Feng Zhao¹, Scott D. Swanson², Jon-Fredrik Nielsen¹, Jeffrey A. Fessler³, and Douglas C. Noll^1
1Biomedical Engineering, The University of Michigan, Ann Arbor, MI, United States, ²Radiology, The University of Michigan, Ann Arbor, MI, United States, ³EECS, University of Michigan, Ann Arbor, MI, United States

We propose to do simultaneous fat saturation and magnetization transfer (MT) preparation with small-tip fast recovery imaging (STFR) which produces high SNR bSSFP-like images free of banding artifacts. This sequence requires much less SAR for MT preparation due to its much higher sensitivity to MT effect than regular non-steady-state MT sequence.

Mauro Costagli¹, Douglas A.C. Kelley², Riccardo Stara³, Gianluigi Tiberi¹, Mirco Cosottini^1,3, and Michela Tosetti^1,4
1IMAGO7, Calambrone, Pisa, Italy, ²GE Healthcare Technologies, San Francisco, California, United States, ³University of Pisa, Pisa, Pisa, Italy, ⁴IRCCS Stella Maris, Pisa, Pisa, Italy

This study presents an IR sequence that enhances the border between two tissues of interest by employing an inversion time such that magnetizations of the two tissues have same magnitude and opposite sign, hence their interface results nulled and highlighted in the image by a dark line. The most obvious advantage of this technique is that it allows for immediate, enhanced visualization of borders between two tissues of interest without any additional postprocessing procedure, hence images are immediately visible to the neuroradiologist in real time. This technique is feasible on both standard and ultra-high field MRI systems.

Ethan M. Johnson¹, Jinyi Qi², Urvi Vyas¹, Kim R. Butts Pauly¹, and John M. Pauly^1
1Stanford University, Stanford, CA, United States, ²UC Davis, Davis, CA, United States

Pulse sequences using very short echo times (‘UTE’, ‘CUTE’, ‘DUTE’, etc.) have been applied variously to image tissues with short spin-spin relaxation (T₂) times. With consideration of the consequences short-T₂ times have for excitation and encoding, a UTE-style pulse sequence capable of visualising bone and other fast-relaxing structures is developed. The acquired images show features that facilitate tissue classification, which is useful, e.g., for MR-PET and high-frequency focussed-ultrasound. To demonstrate, bone, soft tissue and air are classified from 1/T₂ and signal level, from which a linear attenuation coefficient map useful for PET reconstruction is generated.

Yiu-Cho Chung¹, Yanjie Zhu¹, Xin Liu¹, and Chao Zou^1
1Paul C. Lauterbur Research Center for Biomedical Imaging, Shenzhen Key Laboratory for MRI, Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen, Guangdong, China

TSE based FLAIR at 7T is challenging due to B1 transmit field inhomogeneity and SAR limitation. We propose inversion recovery prepared PSIF (IR-PSIF) for FLAIR imaging at 7T. The short TR in PSIF shortens scan time (~5s per slice) and eliminates the need for segmentation. Use of excitation pulses instead of refocusing pulses makes PSIF less sensitive to hardware imperfections. The SNR of white matter from an IR-PSIF image is about 44, and is half that of FLAIR. Such SNR would be sufficient for diagnostic purposes. We believe IR-PSIF may be a fast and robust alternative for FLAIR at 7T.

Ovidiu C. Andronesi¹, Himanshu Bhat², Shreya Mukherjee¹, Peter Caravan¹, and Bruce R. Rosen^1
1Martinos Center, Radiology, Massachusetts General Hospital, Charlestown, Massachusetts, United States, ²Siemens Medical Solutions, Charlestown, Massachusetts, United States

T1 and T2 relaxation in the rotating frame (T1rho, T2rho) are sensitive to slow molecular dynamics on the ms time scale, relevant for interaction of water with important biological systems such as proteins, glycosaminoglycans and cell membranes. T1rho and T2rho have been used to investigate pathology of stroke, Alzheimer’s Disease, Parkinson Disease, liver cirrhosis and cartilage damage. Because T1rho and T2rho contrasts are created by the application of a long spin lock RF field, or by a train of adiabatic inversion pulses this results in sequences that have high specific absorption rate (SAR). Here we propose a pulse sequence based on low power gradient offset independent adiabatic pulses (GOIA-W(n,m))for decreasing SAR and performing slice selective T1rho and T2rho mapping.

Eric K. Gibbons¹, Manoj Saranathan², Brian K. Rutt², John M. Pauly³, and Adam B. Kerr^3
1Department of Bioengineering, Stanford University, Stanford, California, United States, ²Department of Radiology, Stanford University, Stanford, California, United States,³Department of Electrical Engineering, Stanford University, Stanford, California, United States

Brain lesion detection at 7T is complicated by the T1 lengthening observed for white and grey matter without a corresponding increase in CSF T1. As a consequence, there is lower SNR and contrast for grey and white matter in FLAIR sequences designed to null CSF signal. The recently developed MP-FLAIR 3D FSE sequence introduced a magnetization preparation sequence to initially saturate white/grey matter Mz so as to allow for greater recovery during the FLAIR inversion period. In this preliminary work, we examine the possibility of eliminating the inversion preparation entirely, using instead a T2 and diffusion preparation sequence, with the goal being CSF signal suppression due to it’s higher diffusion, and simultaneous introduction of combined T2/diffusion contrast between white/grey matter and lesions.

Steffen Volz¹, Ulrike Nöth¹, Alina Jurcoane², Ulf Ziemann³, Elke Hattingen², and Ralf Deichmann^1
1Brain Imaging Center, University Frankfurt, Frankfurt, Germany, ²Department of Neuroradiology, University Frankfurt, Frankfurt, Germany, ³Department of Neurology, University Frankfurt, Frankfurt, Germany

Correcting for the receiver profile (RP) of the RF coil is a crucial step in quantitative proton density (PD) mapping. Two new RP correction methods were compared for pathological tissue: (1) bias field correction method and (2) pseudo PD calculation method assuming a linear relationship between T1 and PD in white and grey matter. Results were consistent for multiple sclerosis and stroke patients (small lesions). However, differences between the two RP correction methods were found in tumour tissue with results from the pseudo PD method being more plausible.

Hahnsung Kim¹, Dong-Hyun Kim¹, and Jaeseok Park^2
1Yonsei University, Seoul, Korea, ²Korea University, Seoul, Korea

In susceptibility weighted imaging, it is important to remove background phase variations while retaining only local field information. The projection onto dipole field method was recently introduced, employing image-based convolution with ROI masks to decompose a background field within the ROI into one originating from dipoles outside the ROI. However, the image-based convolution is sensitive to rapid field variations at the boundaries of the ROI, leading to loss of important phase information particularly at the air-tissue interfaces. To regionally correct rapid phase aliasing at the boundaries while locally removing smoothly varying background phases, we develop a robust SWI method employing: 1) Single-slab 3D GRASE to simultaneously acquire both artifact-free magnitude images from spin echo and phase images from free induction decay signals, 2) Removal of rapid background field at the boundaries using forward field calculation, and 3) Removal of local background field with the ROI extended to the boundaries to preserve susceptibility induced phase information over the entire image.

Yohan van de Looij^1,2, Rajika Maddage², Nicolas Kunz^1,2, Petra Susan Hüppi¹, Rolf Gruetter^3,4, and Stéphane V. Sizonenko^1
1Division of Child Growth & Development, University of Geneva, Geneva, Switzerland, ²Laboratory for Functional and Metabolic Imaging, Ecole Polytechnique Fédérale de Lausanne, Lausanne, Switzerland, ³Laboratory for Functional and Metabolic Imaging, École Polytechnique Fédérale de Lausanne, Lausanne, Switzerland, ⁴Department of Radiology, University of Geneva and Lausanne, Geneva and Lausanne, Switzerland

DTI gives useful white matter information such as fiber direction and integrity. The phase of gradient echo images (PI) has been used to create anatomical images with excellent grey/white matter contrast. In this work the potential correlations between DTI and PI in the rat brain were investigated. We show that PI contrast is influenced by degree of anisotropy and fiber orientation. A correlation between DTI derived parameters and frequency shift of PI was also observed. Further experiments are in progress to quantify more accurately these results by an assessment of the correlations between DTI and susceptibility maps.

Uzi Eliav¹, Michal E. Komlosh², Peter J. Basser², and Gil Navon^1
1School of Chemistry, Tel Aviv University, Tel Aviv, Israel, ²STBB/PPITS/NICHD/NIH, NIH, Bethesda, MD, United States

The combined double quantum (DQ) filtering and magnetization transfer (DQF-MT) and Ultra-Short TE (UTE) MRI (DQF-MT/UTE) pulse sequence was applied for the study of the annulus fibrosus in porcine intervertebral disc (IVD). It was shown that unlike the UTE that gives the same intensity throughout the tissue the intensity in the DQF-MT/UTE image declines from the outer layers towards the inner ones (closer to the nucleus pulposus). The latter result was interpreted as a combined effect of the declined concentration of collagen and the ratio of collagen type I to type II going from the outer into the inner layers.

Henry H. Ong¹, Michael J. Wilhelm², Suzanne L. Wehrli³, Edward James Delikatny⁴, and Felix W. Wehrli^1
1Laboratory for Structural NMR Imaging, Department of Radiology, Perelman School of Medicine at the University of Pennsylvania, Philadelphia, PA, United States, ²Department of Chemistry, Temple University, Philadelphia, PA, United States, ³NMR Core Facility, Children's Hospital of Philadelphia, Philadelphia, PA, United States, ⁴Department of Radiology, University of Pennsylvania, Philadelphia, PA, United States

Understanding MR properties of myelin would provide valuable insight into current MRI methods for myelin imaging. Myelin forms a lamellar liquid crystal lipid system whose phase behavior is influenced by temperature. In recent work the MR properties of rehydrated myelin extracts have been studied at room temperature, but whether the results apply to physiological temperature is not currently understood. Here, we studied the temperature dependence (10-50°C) of the liquid crystal phase of rehydrated bovine myelin extract with ³¹P and ¹H NMR. The results are consistent with myelin lipids exhibiting a lamellar liquid crystal phase over this temperature range

Hyunyeol Lee¹, Woo Chul Jeong², Hyung Joong Kim², Eung Je Woo², and Jaeseok Park^1
1Brain and Cognitive Engineering, Korea University, Seoul, Korea, ²Biomedical Engineering, Kyung Hee University, Yongin, Gyeonggi, Korea

Magnetic resonance electrical impedance tomography (MREIT) was recently introduced to achieve high spatial resolution, wherein the internal magnetic flux density (Bz) induced by current injection results from image phases and electrical conductivity is then calculated using the harmonic Bz algorithm. To achieve accurate conductivity distribution in tissues, a high signal-to-noise ratio (SNR) in Bz is critical, which is proportional to the product of current injection time (TC) and SNR in magnitude image. To effectively enhance the SNR of Bz in MREIT and speed up data acquisition, in this work we develop a Bz-SNR-optimized echo-shifted incoherent steady state imaging pulse sequence for accurate quantification of electrical conductivity, wherein free induction decay (FID) signals experience multiple current injections to form an echo without apparent loss of signals while retaining high imaging efficiency.

Leeor Alon^1,2, Christopher Collins^1,2, Giuseppe Carluccio¹, Dmitry S. Novikov¹, Yudong Zhu^1,2, and Daniel Sodickson^1,2
1Department of Radiology, Bernard and Irene Schwartz Center for Biomedical Imaging, New York University, New York, NY, United States, ²Sackler Institute of Graduate Biomedical Sciences, New York University, New York, NY, United States

At last year’s ISMRM meeting, we introduced Local Maxwell Tomography (LMT) – a method for noninvasive mapping of electrical conductivity and permittivity using incomplete MR-based measurements of RF field curvature. Here, we apply a theoretical framework analogous to that of LMT to solve Pennes’ bio-heat equation, instead of Maxwell’s equations, and thereby to reconstruct tissue thermal property maps from MR thermometric measurements. We outline the theory and demonstrate reconstruction of thermal properties such as heat capacity, thermal conductivity, and metabolic energy rate.

Xia Jiang¹, Hanbing Lu², Shuichi Shigeno³, Li-Hai Tan⁴, Yihong Yang², Clifton W. Ragsdale³, and Jia-Hong Gao^1
1Brain Research Imaging Center and Department of Radiology, University of Chicago, Chicago, IL, United States, ²Neuroimaging Research Branch, National Institute on Drug Abuse, National Institutes of Health, Baltimore, MD, United States, ³Department of Neurobiology, University of Chicago, Chicago, IL, United States, ⁴State Key Lab for Brain Research, University of Hong Kong, Hong Kong, China

Much effort has been made to determine the feasibility of neuronal current MRI (nc-MRI), where the small magnetic field change due to neuronal currents activity is mapped using MRI. However, most previous studies in nc-MRI were susceptible to the contamination of BOLD signal. In this study we tested nc-MRI in the octopus visual system, which is free from BOLD effect. In addition, electrophysiological recordings were used to confirm electrical activities. Scans with the EPI sequences indicated that no statistically significant nc-MRI effect could be detected at 0.2%/0.2° level for signal magnitude and phase respectively.

Luning Wang¹, Yangqing Lu², Khan Hekmatyar³, Steve Stice², and Qun Zhao^1
1Department of Physics and Astronomy, University of Georgia, Athens, GA, United States, ²Department of Animal and Dairy Science, University of Georgia, Athens, GA, United States, ³BioImaging Research Center (BIRC), University of Georgia, Athens, GA, United States

Magnetic resonance imaging (MRI) has been used for studying the dynamics of chick embryo development and growth. This can be repeated observations on the same embryo on a longitudinal basis, using high resolution ultra-high-field (7 Tesla or higher) MR scanners. MRI-based cell tracking using super-paramagnetic iron oxide (SPIO) particles provide an excellent means of cell monitoring in in-vivo. In this work, a novel sweep imaging with Fourier transformation (SWIFT) sequence, compared with conventional fast spin echo (FSE) and gradient-recalled echo (GRE) sequences, will be used to scan a chick embryo injected with SPIO nanoparticles.

Alan B. McMillan¹, Hyungseok Jang¹, Sankaran Subramanian², and Murali C. Krishna^2
1Radiology, University of Wisconsin, Madison, WI, United States, ²Radiation Biology Branch, Center for Cancer Research, NCI, National Institutes of Health, Bethesda, MD, United States

Electron paramagnetic resonance imaging (EPRI) employed using single-point (SP) imaging techniques offers the capability to dynamically image in vivo tumor oxygenation. While highly specific, current capabilities for high spatial resolution dynamic imaging are limited. Recently, we have investigated partial k-space acceleration techniques for SP-EPRI. In this work, we show that the SP-EPRI imaging technique is highly suited to PF techniques due to its tractable phase characteristics, and that performance of each method is dependent upon image matrix size.

Frank Golub¹, Lee C. Potter¹, and Rizwan Ahmad^1
1The Ohio State University, Columbus, OH, United States

The current standard for evaluating the T2 decay constant in pulsed EPR oximetry employs peak-picking of spin echoes. This method is inefficient, because it only uses a small subset of the available echo data. In this work, we characterize a sequence of measured spin echoes using either a rank-one matrix or a train of parametric functions. Simulation results indicate that the proposed methods can significantly reduce the error variance in the estimation of T2 and hence the data acquisition time for EPR oximetry.

Hyungseok Jang¹, Sankaran Subramanian², Nallathamby Devashayam², Murali C. Krishna², and Alan B. McMillan^1
1Radiology, University of Wisconsin, Madison, WI, United States, ²Radiation Biology Branch, Center for Cancer Research, NCI, National Institutes of Health, Bethesda, MD, United States

In EPRI acquired using single-point techniques, repeated imaging experiments with differing gradient amplitudes are used to calculate spectral linewidths necessary to quantify pO₂. In this study, we propose a novel estimation method that enables direct calculation of linewidth within a single dataset. We investigate gridding reconstruction and image-based registration techniques to maintain image FOV, and enable pixelwise estimation of linewidth. In phantom experiments, we show that these methods allow quantification of pO₂ within a single dataset, which will allow temporal resolution improvements of 3x compared to existing techniques.

Celine M. Desmet¹, Philippe Leveque¹, Gaetane Leloup², and Bernard Gallez^3
1Biomedical Magnetic Resonance, Universite catholique de Louvain, Brussels, Belgium, ²CRIBIO, Universite catholique de Louvain, Brussels, Belgium, ³Biomedical Magnetic Resonance, University of Louvain, Brussels, Belgium

Tooth enamel associated with EPR detection is a very promising natural dosimeter. When conducting retrospective dosimetry, attention should be paid to recent restorations on teeth (less than 6 month), specially for doses lower than 3Gy because dental resins exhibit intense EPR signals early after photopolymerization

Ge Zhang¹, Xianlong Wang¹, Shilong Lu¹, Jinyuan Zhou², and Zhibo Wen^1
1Department of Radiology, Zhujiang Hospital, Southern Medical University, Guangzhou, Guangdong, China, ²Department of Radiology and Radiological Science, Johns Hopkins University School of Medicine, Baltimore, Maryland, United States

We explored the imaging features of different grades of glioma using Amide Proton Transfer MR Imaging at 3 Tesla. Results showed that the average APT signal intensities of the viable tumor cores were significantly higher than those of peritumoral edema or normal-appearing white matter in both low-grade and high-grade glioma. The average APT signal intensities of the viable tumor cores were significantly higher than in low-grade glioma. APT imaging provides additional diagnostic information to characterize different grade of glioma.

Jinyuan Zhou^1,2 and Xiaohua Hong^1
1Department of Radiology, Johns Hopkins University, Baltimore, Maryland, United States, ²F.M. Kirby Research Center for Functional Brain Imaging, Kennedy Krieger Institute, Baltimore, Maryland, United States

We performed APT imaging and NOE imaging experiments on 9L tumor-bearing rats. The NOE effect is clearly visible at lower saturation powers and is larger in contralateral normal brain tissue. On the contrary, the APT effect is observed at relatively higher saturation powers and is larger in tumor. The NOE effect is a confounding factor for the quantification of APT and other CEST effects and may be exploited as a potentially new MR contrast.

Tao Jin¹ and Seong-Gi Kim^1,2
1Department of Radiology, University of Pittsburgh, Pittsburgh, PA, United States, ²Department of Neurobiology, University of Pittsburgh, Pittsburgh, PA, United States

Recent in vivo saturation transfer MRI studies have showed significant signals from the Nuclear Overhauser effect (NOE) from aliphatic protons of macromolecules. In this preliminary study, we found that the aromatic protons, with frequency range downfield of water and close to the amide frequencies, can also affect the water signal in a saturation transfer experiment through the NOE, and the effect is significant in vivo. Both aromatic and aliphatic NOE signals are not sensitive to tissue pH, while they impose difficulties to the quantification of amide proton transfer contrast, they may provide novel MRI contrast complementary to CEST.

Christian C. Mirkes^1,2, Jens Hoffmann¹, Gunamony Shajan¹, Rolf Pohmann¹, and Klaus Scheffler^1,2
1High-Field MR Center, Max Planck Institute for Biological Cybernetics, Tübingen, Germany, ²Department for Biomedical Magnetic Resonance, University of Tübingen, Tübingen, Germany

Amide proton transfer imaging is very promising at ultra-high field due to the large spectral separation. Here we show that quantification of the chemical exchange saturation effect can be facilitated by homogenizing the transmit field through B1 shimming.

Xiaopeng Zong¹, Ping Wang¹, Seong-Gi Kim¹, and Tao Jin^1
1Department of Radiology, University of Pittsburgh, Pittsburgh, PA, United States

Amide-water proton transfer (APT) can be exploited to provide useful MRI contrast under various disease conditions. However, the APT-weighted signals (APTw) might be contaminated by other chemical exchange processes, such as fast amine-water proton exchange. To investigate this issue, we measured APTw of nine major neurometabolites in phantoms and estimated their contributions to the in vivo APTw contrast in ischemic rat brain. We found significant contributions from several metabolites to the APTw, which showed pH dependence opposite to the APTw signal in vivo. Thus, possible contaminations from neurometabolites should be considered when interpreting the sources of APTw contrast.

Guang Jia¹, Saba N. Elias¹, Ronney Abaza², Debra L. Zynger³, Zarine K. Shah¹, Lai Wei⁴, Robert R. Bahnson², and Michael V. Knopp^1
1Department of Radiology, The Ohio State University, Columbus, OH, United States, ²Department of Urology, The Ohio State University, Columbus, OH, United States,³Department of Pathology, The Ohio State University, Columbus, OH, United States, ⁴Center for Biostatistics, The Ohio State University, Columbus, OH, United States

This study shows the importance of APT-MRI contrast optimization based on pulse duration and power strength in clinical prostate cancer patient examinations and its benefits to further develop this methodology into an imaging biomarker.

Jochen Keupp¹, Guang Jia², Ivan E. Dimitrov³, Silke Hey⁴, and Michael V. Knopp^2
1Philips Research, Hamburg, Germany, ²Department of Radiology, The Ohio State University, Columbus, OH, United States, ³Philips Healthcare, Cleveland, OH, United States, ⁴Philips Healthcare, Best, Netherlands

Endogenous contrast of exchangeable amide protons of intra-cellular proteins and peptides (amide proton transfer, APT) has been developed sucessfully and applied for first cinical studies in oncology applications. The work up to now has been mostly focused on the brain, with some extensions to e.g. prostate and breast APT. The organs addressed so far are not subject to respiratory motion. Body oncology applications in e.g. liver and kidney are challenging for APT, because the time scale of RF saturation needed for sensitive APT MRI (about 2s) is of the same order as typical respiratory intervals. A basic respiratory triggered saturation transfer technique was previously developed for contrast agent studies in human kidneys. In the current work, this approach was extended for improved stability of the contrast with varying respiratory cycles, scan time efficiency and SAR management. The novel technique was successfully tested in human volunteers with a focus on APT contrast in liver and kidney.

Xue Yu¹, Elaine Yuen Phin Lee¹, Queenie Chan², and Mina Kim^1
1Diagnostic Radiology, The University of Hong Kong, Hong Kong, China, ²Philips Healthcare, Hong Kong, China

Neoadjuvant treatment for peritoneal metastasis has emerged. Therefore evaluation of peritoneal metastasis is crucial for diagnosis and treatment planning. As APT MRI can provide complementary information to current MR sequences in clinical use, the purpose of this study is to assess APT MRI in peritoneal metastasis imaging. This pilot study demonstrates the feasibility of APT MRI in differentiating peritoneal metastasis from muscle and fat, which is clinically relevant as peritoneal metastases can be deposited in the area with abundant fat or adherent to soft tissue. The separation of these tissues will aid detection and assessment of peritoneal metastasis.

Rachel Scheidegger^1,2 and David C. Alsop^1,3
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 spin-locking RF preparations schemes optimized to generate chemical exchange saturation transfer (CEST) contrast with reduced errors from multiple exchangeable protons and intrinsic magnetization transfer (MT) asymmetry for human imaging at 3T. Tuning both the spin-locking power and duration can be used to maximize signal from either amide or amine protons. We demonstrate, in healthy volunteers, how these methods allow clear and robust measurement of the amine proton peak with reduced MT asymmetry which could improve the feasibility of quantifying exchange rates in-vivo and measuring pH.

Guillaume Duhamel¹, Olivier M. Girard¹, Gopal Varma², Patrick J. Cozzone¹, Virginie Callot¹, and David C. Alsop^2
1CRMBM / CNRS 7339, Aix-Marseille University, Marseille, France, ²Department of Radiology, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, MA, United States

Specific imaging able to provide quantification of myelin concentration would be a very valuable tool for clinical and preclinical studies of white matter pathologies. While several advanced MR techniques are used to assess myelin content (short T2 imaging, MWF, MT …), they all are affected by confounding factors which limit their specificity to myelin. A new MT approach specific to the inhomogenous component of the MT spectrum was applied on human at 3T and appeared to be selectively sensitive to tissue with myelin. Here, we investigated the specificity of this inhomogenous MT sequence in mouse brain at 11.75T

Gopal Varma¹, Fotini Kourtelidis¹, and David C. Alsop^1
1Radiology, Division of MR Research, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, United States

Application of quantitative magnetization transfer (MT) to white matter in the brain is of interest, particularly in studies of degenerative diseases. A greater contrast and sensitivity has been shown from inhomogeneous MT (IHMT) imaging. A more quantitative analysis of the IHMT effect is applied in hair conditioner phantoms and in vivo based on a 2 pool model to elucidate exchange time and T₂ parameters associated with an inhomogeneous component. An exchange time of ~6ms and T₂ of ~200µs is found in healthy volunteers, which are distinct from the results in vitro.

Jeffrey William Barker^1,2, Kyongtae Ty Bae¹, and Sung-Hong Park^1,3
1Radiology, University of Pittsburgh, Pittsburgh, Pennsylvania, United States, ²Bioengineering, University of Pittsburgh, Pittsburgh, Pennsylvania, United States, ³Bio and Brain Engineering, Korean Advanced Institute of Science and Technology, Daejeon, Yuseong-gu, Korea

For our newly developed MT ratio (MTR) imaging, we assessed characteristics of interslice MT signals, that are inherent to multi-slice bSSFP imaging for varying flip angle and phase encoding (PE) order. We compared in vivo MTR values to two-pool model simulations. Overall, MTR and SNR values increased with flip angle. Centric PE images showed higher MTR values, higher SNR, and better WM/GM contrast. Linear PE image contrast was influenced by relaxation effects. Simulations agreed well with in vivo data. Our study assessed characteristic factors of the interslice MTR method to be considered for MTR imaging in the brain.

Rachel Scheidegger^1,2, Eric T. Wong^3,4, and David C. Alsop^1,5
1Radiology, Beth Israel Deaconess Medical Center, Boston, MA, United States, ²Health Sciences and Technology, Harvard-MIT, Cambridge, Massachusetts, United States,³Brain Tumor Center & Neuro-Oncology Unit, Beth Israel Deaconess Medical Center, Boston, MA, United States, ⁴Neurology, Harvard Medical School, Boston, MA, United States, ⁵Radiology, Harvard Medical School, Boston, MA, United States

We report a study targeted at separating the different saturation transfer signals from amide, amine, aliphatic protons and macromolecular magnetization transfer contrast (MTC) and identifying their relative contributions to CEST contrast in high grade glioma patients. Amide exchange could be detected with lower saturation power than has previously been reported in glioma but showed no detectable contrast in tumors. At high saturation powers, amine proton exchange was a major contributor to the observed signal but also showed no contrast in tumors. The loss of broad macromolecular MTC from normal brain tissue was responsible for the majority of contrast with glioma.

Richard D. Dortch^1,2, Adrienne N. Dula^1,2, Francesca Bagnato^1,2, David R. Pennell^1,2, Siddharama Pawate³, Simon Hametner⁴, Hans Lassmann⁴, John C. Gore^1,2, and Seth A. Smith^1,2
1Radiology and Radiological Sciences, Vanderbilt University, Nashville, TN, United States, ²Vanderbilt University Institute of Imaging Science, Vanderbilt University, Nashville, TN, United States, ³Neurology, Vanderbilt University, Nashville, TN, United States, ⁴Center for Brain Research, Medical University, Vienna, Austria

We recently developed a selective inversion recovery (SIR) quantitative magnetization transfer (qMT) protocol that exploits the increased signal-to-noise ratio available at 7.0 T to reduce scan times. The goals of this study were: 1) to establish the relationship between the resulting SIR-derived metrics and pathological changes in relapsing-remitting multiple sclerosis (RRMS) patients and 2) to validate the SIR technique by comparing qMT parameter maps in postmortem brains to histological measurements of myelin. The in vivo results suggest that SIR-derived metrics are sensitive to RRMS pathology, and the postmortem results suggest that the method is reporting on changes in myelin content.

Rebecca S. Samson¹, Nils Muhlert², Varun Sethi², Claudia Angela M. Wheeler-Kingshott², Maria A. Ron², David H. Miller², and Declan T. Chard^2
1NMR Research Unit, Queen Square MS Centre, Department of Neuroinflammation, UCL Institute of Neurology, London, England, United Kingdom, ²NMR Research Unit, Queen Square MS Centre, Department of Neuroinflammation, UCL Institute of Neurology, London, United Kingdom

Histopathology has demonstrated extensive cortical grey matter (CGM) demyelination in multiple sclerosis (MS), and suggests that sulcal folds may be preferentially affected, particularly in secondary progressive (SP) MS. We measured sulcal and gyral crown magnetisation transfer ratio (MTR) in MS and healthy control CGM in vivo, and examined associations with clinical status. CGM MTR abnormalities were present in all MS subtypes and were most pronounced in SPMS. However, sulcal was lower than gyral MTR in all groups (including controls) except SPMS, indicating that there is not a clear sulcal predilection for these changes.

Kejia Cai¹, Mohammad Haris¹, Anup Singh¹, Ravi Prakash Reddy Nanga¹, Ranjit Ittyerah¹, Damodar Reddy¹, Harish Poptani¹, Hari Hariharan¹, and Ravinder Reddy^1
1University of Pennsylvania, Philadelphia, PA, United States

The CEST effect at 2ppm (CEST@2ppm), quantified through fitting Z spectrum with Lorentzian functions, decreases in tumor compared to normal brain tissue and further reduces as tumor progresses, correlating to MRS quantification of creatine. Along with other justifications, we suggest creatine to be the major contribution to the CEST@2ppm. Our results show F98 tumors have significant lower CEST@2ppm effect compared to 9L tumors. Once validated, the high-resolution creatine mapping method could be used for grading brain tumors.

Tao Jin¹ and Seong-Gi Kim^1
1Department of Radiology, University of Pittsburgh, Pittsburgh, PA, United States

CEST imaging utilizing the the endogenous amide-proton transfer (APT) effect has shown great potential in stroke studies. However, the APT contrast in stroke is negative which decreases at the lesion region. A positive pH-sensitive imaging contrast would be preferable. In this preliminary study, we investigated the chemical exchange effects of endogenous guanidine and hydroxyl protons and their potential application in stroke studies. Because guanidine and hydroxyl protons exchange with water at much faster rate than amide protons, a positive chemical exchange imaging contrast induced by tissue acidosis can be detected by judicious selection of the off-resonance irradiation pulse power.

Catherine DeBrosse¹, Feliks Kogan¹, Mohammad Haris¹, Anup Singh¹, Kejia Cai¹, Ravi Prakash Reddy Nanga¹, Hari Hariharan¹, and Ravinder Reddy^1
1Center for Magnetic Resonance and Optical Imaging, Department of Radiology, University of Pennsylvania, Philadelphia, Pennsylvania, United States

The purpose of this study was to determine the feasibility of mapping creatine (Cr) in vivo using chemical exchange saturation transfer (CEST) on routine clinical scanners (3T). Mild plantar flexion exercise on healthy human subjects led to an increase in CrCESTasym that is correlated to the change in the PCr signal measured with 31P MRS.

Nirbhay N. Yadav^1,2, Craig K. Jones^1,2, Jun Hua^1,2, Jiadi Xu^1,2, and Peter C.M. van Zijl^1,2
1Russell H. Morgan Department of Radiology and Radiological Science, Johns Hopkins University School of Medicine, Baltimore, MD, United States, ²FM Kirby Research Center, Kennedy Krieger Institute, Baltimore, MD, United States

It is shown that it is possible to image endogenous exchangeable proton signals in the human brain at 3T using the recently reported method of frequency labeled exchange transfer (FLEX) MRI. The results indicate that, compared to traditional CEST experiments, FLEX MRI preferentially detects more rapidly exchanging amide/amine protons, thereby changing the information content of the exchangeable proton spectrum. This has the potential to open up different types of endogenous applications as well as more easy detection of rapidly exchanging protons in diaCEST agents or fast exchanging units such as water molecules in paraCEST agents without interference of conventional MTC.

Xiaolei Song^1,2, Jiadi Xu^1,2, Shuli Xia³, Nirbhay N. Yadav^1,2, Bachchu Lal³, Jeff W.M. Bulte^2,4, John Laterra³, Peter C.M. van Zijl^1,2, and Michael T. McMahon^1,2
1Division of MR Research, The Russell H. Morgan Department of Radiology and Radiological Science, The Johns Hopkins University, Baltimore, Maryland, United States, ²F.M. Kirby Research Center for Functional Brain Imaging, Kennedy Krieger Institute, Baltimore, MD, United States, ³Department of Neuro-Oncology, Kennedy Krieger Institute, Baltimore, MD, United States, ⁴Division of MR Research, The Russell H. Morgan Department of Radiology and Radiological Science, Johns Hopkins University, Baltimore, Maryland, United States

Chemical Exchange Saturation Transfer (CEST) has become an important method to generate MRI contrast, with applications in the imaging of strokes and cancer. We present an efficient CEST imaging method, named Multi-echo Length and Offset VARied Saturation (Me-LOVARS), which allows acquiring a series of Z-spectra at multiple t_sat values, without extra scan time cost. When using Me-LOVARS for imaging mice baring glioblastomas, 3 Z-spectra with saturation length of 0.8sec., 1.6sec. and 2.4sec were acquired, using same time as one conventional Zspectrum. The contrast-noise-ratio of B₀-corrected MTRasym maps using Me-LOVARS is also comparable to the conventional method with the same t_sat.

Volkert Brar Roeloffs¹, Moritz Zaiss¹, and Peter Bachert^1
1Medical Physics in Radiology, German Cancer Research Center, Heidelberg, Baden-Wuerttemberg, Germany

Chemical exchange saturation transfer (CEST) experiments provide a NMR-imaging technique capable of detecting dilute metabolite protons. Off-resonant spinlock experiments are in a similar way sensitive to chemical exchange. However, within the clinical context, RF pulse duration and power are limited due to SAR-restrictions. For this reason, in this work the effect of pulsed spinlock saturation has been modeled theoretically. The findings were validated experimentally and a method of quantifying exchange rates was proposed, which might be an important step towards a contrast-agent free imaging technique that allows in vivo quantification of physiological relevant exchange rates and metabolite concentrations.

Sheng-Min Huang¹, Chih-Kuang Yeh¹, and Fu-Nien Wang^1
1Department of Biomedical Engineering & Environmental Sciences, National Tsing Hua University, Hsinchu, Taiwan

The interference from magnetization transfer (MT) in chemical exchange saturation transfer (CEST) complicates the issue of quantitative measurement. In this study, we performed both numerical simulation and phantom experiment to identify if linear combining assumption of CEST and MT signal is appropriate. Both our results showed that the mix signal of CEST and MT were overestimated approximately 25% by linear model, which indicates that linear combination fails to describe the relations between CEST and MT signal. Comprehensive study of CEST and MT relationship is needed for reducing the asymmetric MT effect.

Moritz Zaiss¹, Steffen Goerke¹, and Peter Bachert^1
1Dept. of Medical Physics in Radiology, German Cancer Research Center (DKFZ), Heidelberg, BW, Germany

CEST evaluation suffers from concomitant effects as direct water saturation (spillover) and MT. We present a new spillover correction for pulsed CEST which yields the exchange dependent relaxation in the rotating frame Rex. We proof our method at creatine model solutions and show additionally, that also quantification is possible and therefore calculation of pH and creatine concentration maps. Our method can simply be applied to clinical data of gagCEST, APT, glucoCEST and many more.

Phillip Zhe Sun¹, Gang Xiao², and Renhua Wu^3
1Radiology, Martinos Center for Biomedical Imaging, Charlestown, MA, United States, ²Department of Math and Information Technology, Hanshan Normal University, Chaozhou, Guangdong, China, ³Department of Medical Imaging, Medical College of Shantou University, Chaozhou, Guangdong, China

Whereas CEST-weighted MRI has provided tremendous insights, it is important to develop quantitative CEST (qCEST) analysis to characterize the underlying CEST system. We here postulated that the concomitant RF spillover effects can be reasonably estimated for simplified qCEST analysis. Specifically, both the labile proton ratio and exchange rate can be determined from RF spillover factor corrected modified omega plot. The proposed qCEST analysis was validated both numerically and experimentally. In summary, our study established a simplified qCEST analysis algorithm, which remains promising to aid the ongoing development of qCEST MRI.

Tao Jin¹ and Seong-Gi Kim^1
1Department of Radiology, University of Pittsburgh, Pittsburgh, PA, United States

In CEST studies, the off-resonance irradiation power amplitude (B₁) should be adjusted to optimize the sensitivity and/or specificity of chemical exchange contrast. If only one type of labile proton exist or is dominant, B₁ can be optimized to maximize the sensitivity of chemical exchange contrast. If several types of labile protons have similar Larmor frequencies but very different exchange rates, B₁ can also be optimized to selectively enhance the chemical exchange contrast from a specific labile proton. In this work we derived analytical solutions for B₁ optimization of these two cases in the slow exchange regime.

Kenya Murase¹, Junpei Ueda¹, Koji Itagaki¹, Shigeyoshi Saito¹, and Atsuomi Kimura^1
1Medical Physics and Engineering, Osaka University, Suita, Osaka, Japan

Recently, there have been an increasing number of studies that have used the chemical exchange effect to probe the tissue microenvironment. Most of these studies adopted either a chemical exchange saturation transfer (CEST) or a spin-locking (SL) approach. We presented a simple and fast method for solving the time-dependent Bloch equations in CEST MRI with SL, and proposed a new method for quantifying chemical exchange rates from CEST MRI with and without SL using this method. Our results suggest that the CEST MRI with SL is more reliable than that without SL for quantifying chemical exchange rates.

Jing Yuan¹, Qinwei Zhang¹, Anil T. Ahuja¹, and Yi-Xiang Wang^1
1Department of Imaging and Interventional Radiology, Chinese University of Hong Kong, Shatin, NT, Hong Kong

Spin-lock technique, which is traditionally used to measure T1£l relaxation time, can also be used for chemical exchange (CE) imaging. This study theoretically investigates the optimization of spin-lock frequency (FSL) and spin-lock time (TSL) to maximize the asymmetric magnetization transfer ratio for different proton exchange rates, by numerical simulations based on a two-pool R1£l relaxation model derived from the Bloch-McConnell equations. Results show that spin-lock pulse with low FSL and long TSL is sensitive to detect slow CE processes while spin-lock pulse with high FSL and short TSL is useful for fast CE process study.

Olivier E. Mougin¹ and Penelope A. Gowland^2
1SPMMRC, School of Physics and Astronomy, University of Nottingham, Nottingham, Select, United Kingdom, ²SPMMRC, University of Nottingham, Nottingham, Nottinghamshire, United Kingdom

CEST is increasingly used to study a wide range of endogenous and exogenous species. However different species of interest overlap in the z-spectra even at high field due to line broadening caused by exchange. Here we propose a new combination of acquisition and analysis approaches to separate fast and slow exchanging species. Simulation and experimentation of a z-spectra database for two pools exchanging at different rates is presented here, and was used successfully in fitting pool protons overlapping on the z-spectra, both numerically and empirically.

Feliks Kogan¹, Randall B. Stafford², Mohammad Haris¹, Erin K. Englund², Anup Singh¹, Kejia Cai¹, Catherine DeBrosse¹, Ravi Prakash Reddy Nanga¹, Hari Hariharan¹, John A. Detre³, and Ravinder Reddy^1
1Center for Magnetic Resonance and Optical Imaging (CMROI), University of Pennsylvania, Philadelphia, PA, United States, ²Department of Radiology, University of Pennsylvania, Philadelphia, PA, United States, ³Center for Functional Neuroimaging, University of Pennsylvania, Philadelphia, PA, United States

The objective of this work was to determine the contribution of perfused blood on the chemical exchange saturation transfer (CEST) from creatine. Perfusion changes were produced by a combination of cuff inflation and deflation. Reactive hyperemia resulted in substantial increases in perfusion as measured by arterial spin labeling (ASL) mean percent difference maps but negligible changes in CrCEST maps. This demonstrates that the CEST effect from perfused blood is negligible and does not confound CrCEST measurements.

Jiadi Xu¹, Nirbhay N. Yadav¹, Amnon Bar-Shir², Craig Jones^1,2, Kannie W.Y. Chan¹, Jingyang Zhang², Piotr Walczak², Michael T. McMahon¹, and Peter C.M. van Zijl^1
1F. M. Kirby Center, Kennedy Krieger Institute, Baltimore, MD, United States, ²Russell H. Morgan Department of Radiology and Radiological Science, Johns Hopkins University, Baltimore, MD, United States

We propose a new technique, variable delay pulsed CEST (VDP CEST), which eliminates the need for performing asymmetry analysis in CEST imaging. The scheme involves acquiring images with variable delay times between RF saturation pulses in pulsed CEST, producing a series of CEST images sensitive to the speed of saturation transfer. Subtracting two images or fitting a time series produces CEST and NOE-relayed CEST maps without effects of direct saturation and strongly reduced influence of semi-solid magnetization transfer. This technique, which can provide very short scan times when using a single RF frequency, is demonstrated in vivo on rat brain.