Min-Oh Kim¹, Joonsung Lee¹, Sang-Young Zho¹, and Dong-Hyun Kim^1
1Electrical and Electronic Engineering, Yonsei University, Seoul, Korea

Aliasings occurred due to undersampled phase-encodings in accelerated MR imaging are separated to readout direction by applying additional gradient along undersampled direction at readout timing. Resolution and FOV are redefined due to voxel modification as sheared form. With this method, 6x acceleration is achieved in high-resolution invivo 3D imaging with a partial loss of diagonal resolution. Further acceleration is possible by using parallel imaging techniques such as SENSE and GRAPPA.

Hans Weber¹, Daniel Gallichan¹, Anna M. Welz¹, Chris A. Cocosco¹, Sebastian Littin¹, Jürgen Hennig¹, and Maxim Zaitsev^1
1Department of Radiology, Medical Physics, University Medical Center Freiburg, Freiburg, Germany

ExLoc allows excitation and geometrically matched spatial encoding of curved slices by the application of a set of nonlinear, but locally orthogonal, encoding fields. In this work we explore ExLoc’s potential for multi-slice imaging. Stacks of curved slices result in improved relevant volume coverage for fewer excited slices and thus increase efficiency for particular applications. We present the results of experiments performed on a phantom, as well as a demonstration in vivo.

Hirad Karimi¹, William Dominguez-Viqueira², and Charles H. Cunningham^1,2
1Dept. of Medical Biophysics, University of Toronto, Toronto, Ontario, Canada, ²Imaging Research, Sunnybrook Research Institue, Toronto, Ontario, Canada

Conventional linear gradient fields generated using gradient coils usually perform spatial encoding in MRI. These coils are stationary relative to any physiological movements and this ultimately limits the spatial resolution. Some applications such as intravascular MR imaging (iMRI) could benefit from the high-resolution images that could result from new approaches. Here, a new encoding technique that allows encoding fields to move with physiological structures is investigated. The method uses only the field perturbations emanating from markers with different susceptibilities as the encoding fields. (The gradient coils on scanner are not used). Simulations and phantom studies were conducted to further validate the result.

Dominique M. R. Corteville¹, Friedrich Wetterling¹, and Lothar R. Schad^1
1Computer Assisted Clinical Medicine, Heidelberg University, Mannheim, Germany

Traditionally RF-encoding is performed by modulating the magnetization along the longitudinal axis of the magnet and rotating it into the orthogonal plane. Recent hardware improvements made it possible to achieve encoding by direct usage of the excitation pulses. This study evaluates the possibilities and challenges of RF-encoding using a single transceiver surface coil, providing general information about the achievable speed and resolution. Sodium was chosen as imaging nucleus to outline the advantages of using nuclei with short relaxation times.

Stephan Orzada^1,2, Sören Johst^1,2, Andreas K. Bitz^1,2, Oliver Kraff^1,2, Mark E. Ladd^1,2, and Stefan Maderwald^1
1Erwin L. Hahn Institute for MRI, Essen, NRW, Germany, ²Department of Diagnostic and Interventional Radiology and Neuroradiology, University Hospital Essen, Essen, NRW, Germany

At field strengths of 7 Tesla and above, severe B1 inhomogeneities cause problems due to non-uniform contrast and SNR distributions. Most methods proposed to mitigate these problems rely on multi-channel transmit systems. In this single-channel variant of TIAMO, we propose acquiring two images with different transmit amplitudes in an interleaved fashion. In this way the dark areas where the refocusing pulse does not achieve flip angles close to α = 180° + (n∙360°) are in different positions in the two images. The resulting combined images show less impact of the inhomogeneous transmit field.

Tiejun Zhao¹, Hai Zheng², Yongxian Qian², Tamer S Ibrahim², and Fernando Boada^2
1Siemens Healthcare USA; Siemens Medical Solutions USA, Inc., Pittsburgh, PA, United States, ²University of Pittsburgh, Pittsburgh, PA, United States

Ultra-short echo imaging in the order of 100us achievable for clinical scanners allowed the detection of protons exhibiting very short T2 relaxation times, which is relevant for tendons, ligaments, or the periosteum. In this abstract, we proposed to use the Binomial pulse excite only the short T2 component and demonstrated that the new scheme helped for detecting fast relaxation tissues with the lower SNR that might be otherwise buried under various imaging artifacts.

An Thanh Vu^1,2, Audrey Chang^1,2, Liyong Chen^1,2, and David Feinberg^1,2
1Helen Wills Neuroscience Institute, University of California Berkeley, Berkeley, CA, United States, ²Advanced MRI Technologies, Sebastopol, CA, United States

Simultaneous echo refocused (SIR) EPI has greatly accelerated the acquisition of both fMRI and DTI data sets. However, the effective echo time (TE) of each individual SIR slice can differ by several milliseconds, resulting in slice dependent signal intensity and BOLD contrast. Here, we propose a constant TE version of the SIR method that is able to use TR specific phase correction navigators while maintaining the minimum TE and TR of the original SIR method. SIR with and without constant TE is evaluated in both phantom and human fMRI experiments.

Mayur Narsude^1,2, José Marques^1,2, Wietske van der Zwaag^1,2, Tobias Kober^1,2, and Rolf Gruetter^1,2
1Laboratory for Functional and Metabolic Imaging, Ecole Polytechnique Fédérale de Lausanne, Lausanne, Switzerland, ²Department of Radiology, University of Lausanne, Lausanne, Switzerland

EPI acquisition times can be shortened by partial Fourier acquisition, parallel imaging and sparse data sampling. We present an approach to reduce the EPI volume acquisition time which accelerates in the slice-encode dimension of 3D-EPI enabling an up to 8 fold acceleration with an 8 channel coil at 7T with good image quality. In the PACEUP-3DEPI (ksPACE acqUisition Parallelized 3DEPI) pulse sequence 2 kspace planes are encoded per rf excitation. 96x96x40 matrix volumes were acquired in 0.9s. BOLD sensitivity was tested via resting state networks and found to be increased compared to 2D and 3D-EPI.

Nan-kuei Chen^1
1Brain Imaging and Analysis Center, Duke University, Durham, North Carolina, United States

Every year millions of patients (e.g., children, seriously ill patients and claustrophobic individuals) cannot complete lengthy MRI procedures without sedation or anesthesia, which poses significant risks for serious adverse effects and harm to health. It is highly desirable to design a novel approach to enable millions of patients to complete clinical MRI without the risk of adverse effects associated with sedation / anesthesia. Here we demonstrate that many clinical MRI sequences may be massively accelerated with segmented echo-planar readout, and high-quality data can be reliably obtained by removing the echo-planar related phase errors with a novel phase-cycled reconstruction algorithm.

Lucilio Cordero-Grande¹, and Carlos Alberola-López^1
1University of Valladolid, Valladolid, Castilla y León, Spain

An extension of HARP reconstruction in tagged MR imaging is presented in which a number of stripe patterns greater than the dimension of the space are acquired in order to robustify the reconstruction of the local phase of the image. Reconstruction equations for the estimation of the deformation gradient tensor are presented and combined with a previous contribution which improves the original HARP method by using the windowed Fourier transform to better capture the spatio-spectral distribution of the magnetization pattern. Results on real images deformed by a synthetic pattern show the benefits of the proposed scheme in limiting the interferences in HARP reconstruction.

L. Keith¹, C. Francois², M. Schiebler², S. Reeder^2,3, and F. Korosec^1,2
1Medical Physics, University of Wisconsin - Madison, Madison, WI, United States, ²Radiology, University of Wisconsin - Madison, Madison, WI, United States, ³Biomedical Engineering, University of Wisconsin - Madison, Madison, WI, United States

We have previously introduced a method for acquisition of time-resolved contrast-enhanced MR angiography of the peripheral vasculature using fractional doses of contrast material (<0.1 mmol/kg). Undersampled 3D radial k-space trajectories (VIPR) and HYPR image processing are combined to provide distal time-resolved MRA with high spatial and temporal resolution while maintaining good image quality with contrast material injections as little as 0.025 mmol/kg in patients with known PVD. Using this method, it is possible to decease the dose of gadolinium-based contrast agents for time-resolved, contrast-enhanced peripheral MRA while maintaining high image quality and diagnostic confidence.

Sebastien Bär¹, Robert Borowiak¹, Jan-Bernd Hövener¹, Jürgen Hennig¹, Dominik von Elverfeldt¹, and Jochen Leupold^1
1Dept. of Radiology, Medical Physics, University Medical Center Freiburg, Freiburg, Germany

The gradient echo (GRE) sequence is one of the standard means to image hyperpolarized nuclei. We show Bloch equation based simulations intending to exploit the decaying hyperpolarized signal over as much sequence cycles as possible. Simulations include both RF-spoiled and non-RFspoiled GRE. Parameters under examination are flip angle and the RF spoil increment. Best signal behaviour was obtained for a moderately small flip angle (approx. 10°) and a small spoil increment (approx. 1°).

Sung-Hong Park¹, Tiejun Zhao², Jung-Hwan Kim¹, Fernando E. Boada^1,3, and Kyongtae Ty Bae^1
1Radiology, University of Pittsburgh, Pittsburgh, PA, United States, ²MR Research Support, Siemens Healthcare, Pittsburgh, PA, United States,³Bioengineering, University of Pittsburgh, Pittsburgh, PA, United States

ALADDIN is a new imaging technique that provides interslice perfusion-weighted and MT asymmetry images. We investigated the effects of gradient imperfections on ALADDIN MT asymmetry imaging in this study. Subtraction artifacts were detectable in ALADDIN MT asymmetry images from an agarose phantom but not from a water phantom, in agreement with the theoretical analysis that they were induced by mismatch in MT frequencies associated with readout eddy currents. The artifacts were suppressed by averaging signals over the readout gradient polarities independent of scan parameters. With suppression of the artifacts, ALADDIN signals of human brain were less dependent on scan conditions.

Kun Zhou¹, and Wei Liu^1
1Siemens (Shenzhen) Magnetic Resonance Ltd., Shenzhen, Guangdong, China

A spin-tagging based method is proposed to correct eddy currents induced distortions in DW-EPI images. In this method strips placed by tagging pulses are used to characterize image distortion. Parameters (scaling and shifting factors) describing distortion are derived by comparing strip images of distorted image and reference image (image with b = 0). After that pixel shifting map is calculated using these parameters and used to perform distortion correction.

Dan Xu¹, Joe K. Maier², and Kevin F. King^1
1Applied Science Lab, GE Healthcare, Waukesha, WI, United States, ²MR Engineering, GE Healthcare, Waukesha, WI, United States

An axial plane based high order eddy current correction (HOEC) method was previously proposed to mitigate diffusion gradient direction dependent image distortion often seen in images acquired with the Stejkal-Tanner diffusion sequence, which is known to produce images with higher signal-to-noise ratio than the corresponding dual spin echo images due to generally shorter echo time. In this paper, we show that HOEC correction can be extended to arbitrary scan plane with arbitrary diffusion direction through the application of gradient and polynomial basis rotation matrices.

Ozan Sayin¹, John A. Derbyshire², Liheng Guo¹, and Daniel A. Herzka^1
1Department of Biomedical Engineering, Johns Hopkins School of Medicine, Baltimore, MD, United States, ²Tornado Medical Systems, Toronto, Ontario, Canada

Recent advancements in parallel imaging and/or image reconstruction (e.g. Compressed Sensing) have led to the common employment of nonconventional and irregular phase encode ordering schemes. For balanced steady-state free precession (b-SSFP) MRI, such sequences can be problematic in terms of eddy current artifacts due to rapid jumps in k-space trajectories from one TR to the next. In this work, we explore and characterize a previously-proposed technique, through-slice dephasing, as an efficient technique for eddy current artifact suppression. We demonstrate the utility of the technique in vivo, and further analyze it via b-SSFP simulations.

Ludovic de Rochefort¹, Eddy S.M. Lee², Matteo Polello³, Luc Darrasse¹, Gianni Ferrante³, and Brian K. Rutt^2
1UMR8081, IR4M (Imagerie par Résonance Magnétique Médicale et Multi-modalités), Univ. Paris-Sud, CNRS, Orsay, France, ²Radiology, Stanford University, Stanford, California, United States, ³Stelar s.r.l, Mede, Italy

Delta relaxation-enhanced magnetic resonance (dreMR) imaging is a B0-cycled technology producing contrast from intended targets only. While ÄB can be achieved with an insertable field-cycling magnet, mutual coupling induces eddy currents (EC) that change the main field. An imaging strategy is presented to characterize and compensate for the EC. A multiphase imaging sequence was modified to produce multiple images after a dreMR pulse. Results show that offset caused by EC is spatially homogeneous, proportional to ÄB and decays monoexponentially. Consequently, the EC induced spatial shifts in the frequency encoding direction which could be compensated for using sequence adaptation and post-processing.

Jiancheng Zhuang¹, Zhong-Lin Lu¹, Christine Bouteiller Vidal¹, and Hanna Damasio^1
1University of Southern California, Los Angeles, California, United States

High angular resolution diffusion images are susceptible to distortions caused by eddy currents induced by large diffusion gradients. A new post-acquisition correction algorithm is proposed which does not require any auxiliary reference scans. Image distortion parameters were obtained by image coregistration, performed only between diffusion weighted images with close diffusion gradient orientations. A linear model that describes distortion parameters (translation, scaling, and shear) as a function of diffusion gradient directions was numerically computed to allow individualized distortion correction for every diffusion-weighted image. The method avoids the problematic procedure of cross-correlating images with significantly different contrasts resulting from very different gradient orientations or strengths. Application of the proposed algorithm in high angular resolution diffusion images markedly reduced eddy current distortions, when compared to results obtained with previously published methods.

Yi Wang^1,2, Emilee Minalga^1,2, Allison Payne², Glen Morrell², and Dennis L. Parker^1,2
1Bioengineering, University of Utah, Salt Lake City, UT, United States, ²Utah Center for Advanced Imaging Research, Salt Lake City, UT, United States

bSSFP sequences offer superior signal intensity in a relative short time. However, fat could appear very bright due to its high T2/T1 values. Fat signal suppression or elimination can be helpful to uncover information that might otherwise be obscured by fat, e.g., lesions and blood vessels in breast imaging. Initial work on achieving two-point Dixon fat and water separation in breast using the dual-echo SSFP sequence has been reported. In this work, we simulate the signal behavior of the dual-echo SSFP to assist in choosing the optimal flip angle for in vivo breast imaging in a breast-specific MR guided high intensity focused ultrasound system.

Cheng-Chieh Cheng^1,2, Hing-Chiu Chang¹, Lawrence Panych², Chun-Jung Juan³, Tzu-Cheng Chao⁴, and Hsiao-Wen Chung^1
1Graduate Institute of Biomedical Electronics and Bioinformatics, National Taiwan University, Taipei, Taiwan, ²Department of Radiology, Brigham and Women's Hospital, Boston, Massachusetts, United States, ³Department of Radiology, Tri-Service General Hospital, Taipei, Taiwan, ⁴Institute of Medical Informatics, National Cheng-Kung University, Tainan, Taiwan

Spatial-Spectral RF pulses have been widely adopted for many purposes, especially in the situation when fat-suppression is necessary. In this study, we utilized this idea to simultaneously generate fat-water separated imaging. For comparison, images with fat/water suppression were acquired using a conventional FLASH sequence. Our results showed these simultaneously acquired images have comparable contrast to the FLASH images. Consequently, our proposed method may be beneficial for studies that require dynamic information of both water and fat.

Yongquan Ye¹, Jiani Hu¹, and E.Mark Haacke^1
1Radiology, Wayne State University, Detroit, MI, United States

In this study, a novel RF-based water-fat separation method is introduced, by utilizing the unique frequence response profile of a pair of opposite-phase rectangular pulses.Functioning as simutaneous excitation and selective suppression, the dual-rect pulses do not require extra scan time but come with low SAR level. With the simplicity and robust performance due to the broad passbands and stopbands, our method is especially advantageous for high field fast 3D imaging.

Samir D Sharma¹, Houchun H Hu², and Krishna S Nayak^1
1Electrical Engineering, University of Southern California, Los Angeles, CA, United States, ²Radiology, Children's Hospital Los Angeles, Los Angeles, CA, United States

Chemical shift-encoded water-fat separation techniques are used in both research and clinical settings because they are highly robust to off-resonance. A tradeoff when using these techniques is a longer scan time since data must be collected at multiple echo-times. Previous works have proposed to use either parallel imaging or compressed sensing to shorten the scan time. In this work, we introduce a joint parallel imaging and compressed sensing approach for water-fat separation. The proposed approach is compared to an existing parallel imaging and water-fat separation method. We demonstrate 3.4x 1D acceleration with the proposed approach.

Cong Zhao¹, and Andrew Liang^2
1Siemens ShenZhen Magnetic Resonance, ShenZhen, GuangDong, China, ²School of Engineering and Applied Science, University of Pennsylvania

The work provides a post-processing algorithm classifying water and fat images generated by symetically acquired 2D/3D dixon techniques. It is also extended to work on imaging with non-connected body parts

Bryan T Addeman¹, Abraam S Soliman^2,3, Curtis N Wiens⁴, and Charles A McKenzie^1,4
1Department of Medical Biophysics, University of Western Ontario, London, ON, Canada, ²The Robarts Research Institute, London, ON, Canada,³Biomedical Engineering, University of Western Ontario, London, ON, Canada, ⁴Department of Physics and Astronomy, University of Western Ontario, London, ON, Canada

Escalating interest in the studies of obesity and metabolic disease has created a demand for techniques to measure the regional distribution of adipose tissue. Current standards employ manual segmentation of fat on a single slice, and extrapolation to estimate total fat volumes. We propose a novel automated tissue segmentation technique on images acquired from the entire intra-abdominal cavity within a single breath-hold. Volunteers were imaged in less than 10 minutes, and the images were analyzed in under 2 minutes. Results are similar to those obtained by manual segmentation, but require no manual intervention and are calculated very rapidly.

Stefan Skare¹, and Anders Lilja^1
1Dept of Clinical Neuroscience, Karolinska Institutet, Stockholm, Sweden

A new propeller sequence has been implemented as a complement to existing T2-w and T2 FLAIR techniques. Based on the classical Spin-Echo sequence, the aim was to produce an image contrast before and after Gd administration that it familiar to the radiologists, while overcoming artifacts from flow and motion. Being a T1-w propeller acquisition, off-resonances from the signal-intense fat manifests as a swirly bright structures instead of a simple shift, why high bandwidths should be used, in particular at 3T.

Guobin Li¹, Dominik Paul², Iulius Dragonu¹, Maxim Zaitsev¹, Jürgen Hennig¹, and Kuan Lee^1
1University Medical Center Freiburg, Freiburg, Baden-Württemberg, Germany, ²Siemens Healthcare, Erlangen, Germany

Partial Fourier acquisition has been widely combined with other acceleration techniques, i.e. Parallel imaging, Compressed sensing etc., to reduce MR scanning time. The k-space central data used for phase correction are crucial for the reconstructed image quality. However, these central data are usually acquired in a simple rectangular pattern, which leads to some problems in 3D MRI. A round pattern for partial Fourier acquisition is presented. Its advantages are also shown with in vivo experiments.

Dae-Hun Kang¹, Jun-Young Chung¹, Da-Eun Kim¹, Young-Bo Kim¹, and Zang-Hee Cho^1
1Neuroscience Research Institute, Gachon University of Medicine and Science, Incheon, Korea

Interleaved multi-shot EPI (iEPI) is able to provide the short echo time and the short echo train length, which lead to reduce geometric distortions. iEPI with a minimum intersegment delay was presented for reducing motion artifacts due to a long acquisition time of iEPI and imaging like ssEPI, namely consecutive interleaved EPI (ciEPI). In this paper, we propose the combination between ciEPI and parallel imaging.

Erik B Beall¹, and Mark J Lowe^1
1Radiology, Cleveland Clinic, Cleveland, OH, United States

Head motion is a huge problem in many advanced MRI sequences. Previous reports of microcoil-based active marker motion correction are truly prospective but impose reductions in SNR and time-efficiency in the main sequence due the separate excitation. Anatomic sequences are largely insensitive to motion, but functional, perfusion and diffusion-weighted EPI presents intractable problems if there is more than minimal head motion. EPI in most uses requires a fat saturation pulse, and we present a method to use fat saturation with active markers to remove the SNR and time penalties and demonstrate proof-of-principle in human data.

R. Todd Constable¹, Leo Tam¹, Jason Stockmann¹, and Gigi Galiana^1
1Diagnostic Radiology, Yale University School of Medicine, New Haven, CT, United States

In this work, a novel approach to spatial encoding that uses nonlinear gradients to impose a unique code on each voxel in the image is introduced. The data is readout in a continuous manner as combinations of linear and nonlinear gradients rotate refocusing different regions of the image at different times. This approach may yield advantages in terms of imaging speed, acoustic noise, and reduced gradient dB/dt, all while providing a high SNR fast acquisition strategy.

Nii Okai Addy¹, Holden H Wu^1,2, and Dwight G Nishimura^1
1Electrical Engineering, Stanford University, Stanford, CA, United States, ²Cardiovascular Medicine, Stanford University, Stanford, CA, United States

 

Mo Kadbi¹, Hui Wang¹, Melanie Traughber², Motaz Alshaher¹, Andrea Yancey³, Jens Heidenreich¹, and Amir A Amini^1
1University of Louisville, Louisville, Kentucky, United States, ²Philips healthcare, Cleveland, United States, ³VA hospital, Louisville, Kentucky, United States

Phase contrast MRI is a non-invasive technique to assess cardiovascular blood flow. However, this technique is not accurate in cases where there is atherosclerotic disease and the blood flow is disturbed.. Carotid bifurcation is one of the main sites of atherosclerosis is and a good example of complex and disturbed blood flow due to atypical geometry of this branch site. Therefore, conventional PC MRI at the site of carotid bifuraction suffers from intravoxel dephasing and flow artifacts. In this work, a 3D UTE PC imaging method is designed to measure the blood velocity in carotid bifurcation using a center-out radial trajectory and short TE time compared to standard PC MRI sequences. With 3D UTE-PC combined quantification and visualization of blood flow in the region of interest is possible.

Zhiyong Zhang¹, Yulan Lin¹, Shuhui Cai¹, and Zhong Chen^1
1Department of Electronic Science, Fujian Key Laboratory of Plasma and Magnetic Resonance, Xiamen University, Xiamen, Fujian, China

To eliminate the influences of field inhomogeneity on the conventional 2D correlation spectroscopy (COSY), two schemes via spatially encoded intermolecular multiple-quantum coherences combined with Hadamard encoding and three-dimensional acquisition respectively were proposed to achieve high-resolution 2D COSY in inhomogeneous fields with high acquisition efficiency. Compared to the conventional method, the acquisition time is shortened by orders of magnitude. The new schemes may be useful for the study of chemical and biological materials.

Samantha A. Telfer¹, N.J. Taylor², James Stirling², Ian Simcock², Anwar R. Padhani², Gareth Thomas¹, and T. W. James^1
1Acuitas Medical, Swansea, West Glamorgan, United Kingdom, ²Paul Strickland Scanner Centre, Northwood, Middlesex, United Kingdom

Fine structure analysis (structural spectroscopy of fine textures below the resolution limit of clinical MR imaging) was applied to specially acquired MR data of ex-vivo livers from pigs and oxen to establish their utility as phantoms for characterising chronic liver disease (CLD). Porcine liver exhibits fibrotic structure similar to human CLD whilst ox liver has an appearance similar to healthy human liver. Whole liver lobules “decorated” with a fibrotic network exhibit a distinctly different spectral signature than the underlying vascular structure of the lobules - hence clearly distinguishing between the two livers and supporting use of fine structure analysis.

Guoxi Xie¹, Xiang Feng², Xin Liu², Bensheng Qiu², and Anthony G. Christodoulou^3
1Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen, Guangdong, China, ²Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, ³Department of Electrical and Computer Engineering, University of Illinois at Urbana-Champaign, Urban

The Partial Separability (PS) model allows sparse sampling for fast MRI. It is generally performed by sampling two datasets (navigator data and image data) to estimate the parameters of the model before reconstruction high spatiotemporal resolution MR images. Based on the theory of partially separable functions, the more spatial frequency components in (k-f) space the navigator data covers, the more accurately the PS model will capture object motion. To address this issue, we present a novel sampling method that uses radial sampling trajectories for navigator data sampling and Cartesian sampling trajectories for image data sampling. This covers more spatial frequency components of the navigator data without requiring re-gridding during image reconstruction.

Hao Sun¹, Jeffrey Fessler¹, and Jon Fredrik Nielsen^2
1Electrical Engineering and Computer Science, University of Michigan - Ann Arbor, Ann Arbor, Michigan, United States, ²Biomedical Engineering, University of Michigan - Ann Arbor, Ann Arbor, Michigan, United States

Small-tip fast recovery (STFR) imaging is a recently proposed steady-state sequence that has similar T2/T1 contrast as bSSFP but has the potential to simultaneously remove banding artifacts and transient fluctuations. At the end of each TR, the STFR method tips the magnetization back to the longitudinal axis using a tailored RF pulse. After the tip-up pulse, a gradient spoiler is applied. In this abstract, we derived an analytic expression for the STFR signal and verified the equation with simulations and phantom data. This new signal equation is useful for analyzing STFR, and potentially could be used for quantitative imaging.

Patrick H Le Roux¹, and Brice Fernandez^2
1Applied Science Lab, GE Healthcare, Palaiseau, France, ²Applied Science Lab, GE Healthcare, Munich, Germany

When a train of RF pulses with constant nutation is modulated in phase according to a periodic quadratic law, a refocusing for all the three components of magnetization (Extreme Echo) is generated at the end of the modulation or after two periods if the period is odd. This was first observed experimentally and also demonstrated mathematically . But that demonstration lacks any physical meaning. We propose here a simple geometrical explanation in which the symmetries of the echo to echo rotation plays an essential role. This remark leads to more general phase modulation having the same property.

Jochen Keupp¹, and Holger Eggers^1
1Philips Research, Hamburg, Germany

Amide proton transfer (APT), a powerful technique for molecular imaging of endogenous proteins, is based on an asymmetry analysis of RF saturation frequency offsets acquired around the water resonance, which needs precise B₀ homogeneity correction. While fast spin-echo (FSE) based APT has superior contrast-to-noise ratio, the previously shown multi-echo APT acquisition with intrinsic Dixon-type B₀ mapping/correction is restricted to gradient-echo sequences. We propose a FSE-Dixon APT technique with varied echo time shifts, using iterative Dixon reconstruction across different positive saturation frequency offsets for intrinsic B₀ mapping and correction. Feasibility in the human head is demonstrated using a clinical 3T scanner.

Anup Singh¹, Kejia Cai¹, Mohammad Haris¹, Hari Hariharan¹, and Ravinder Reddy^1
1Radiology, University of Pennsylvania, Philadelphia, Pennsylvania, United States

In current study effect of B1 in-homogeneities on in vivo chemical-exchange-saturation-transfer(CEST) asymmetry contrast from human brain at 7T is studied and a procedure for correction of these in-homogeneities is presented. The z-spectral and CEST data (at 3ppm) at multiple saturation powers and fixed duration(1s) were obtained from brain of healthy human volunteers at 7T research scanner. These data were corrected for B0 in-homogeneities before generating z-spectral and CEST asymmetry contrast. B1 map from human brains were highly in-homogeneous (~50% variation) and CESTasy map without these corrections were difficult to interpret. Proposed calibration based approach resulted in B1-inhomogeneities corrected CESTasy map.

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

One practical issue for the chemical exchange (CE) based MRI techniques is that the conventional T₁ and T₂ relaxations often contaminate the CE imaging contrast. Another issue is that due to hardware or specific absorption rate limitations, CE-sensitive image often has to be acquired using a short irradiation pulse, although a long pulse reaching the steady state may simplify the quantification and sometimes results in better image contrast. Here we propose a novel acquisition method that can use a relatively short irradiation pulse to remove T₁ and T₂ effects in the CE contrasts, and also to obtain the steady-state imaging contrast.

Tae Kim¹, Kristy Hendrich¹, and Seong-Gi Kim^1
1Radiology, University of Pittsburgh, Pittsburgh, PA, United States

We compare various strategies utilizing inversion prepulses applied at water proton frequency and/or off-resonance MT irradiation to obtain magnetization transfer ratio (MTR) maps. Our results demonstrate that MTR maps of similar quality can be obtained at significantly reduced irradiation times (i.e., much lower overall RF power deposition) when incorporating on-resonance inversion prepulses, as compared with steady-state MTR maps acquired with only off-resonance MT irradiation.

Nirbhay Yadav^1,2, Amnon Bar-Shir¹, Craig Jones^1,2, Chien-Yuan Lin^2,3, Jun Hua^1,2, Assaf Gilad¹, Michael McMahon^1,2, and Peter van Zijl^1,2
1Johns Hopkins University, Baltimore, MD, United States, ²KKI, Baltimore, MD, United States, ³UT Southwestern Medical Center, Dallas, TX, United States

In vivo CEST studies are complicated by multiple groups of protons that contribute to the exchange contrast. Here, we used the recently reported FLEX approach to measure the exchange-based proton-transfer ratio (PTR) from the imino protons of thymidine under conditions of different exchange rate due to a change in pH of the solution. Our results show that it is possible to weight exchange contrast in FLEX based on exchange rate. These methods can then be used to filter the contrast from protons with certain exchange rates.

Xiaolei Song^1,2, Amnon Bar-Shir^1,3, Yajie Liang^1,3, Guanshu Liu^1,2, Assaf A Gilad^1,3, Jeff W.M. Bulte^1,3, Peter C.M. van Zijl^1,2, and Michael T Mcmahon^1,2
1Russell H. Morgan Department of Radiology and Radiological Sciences, Johns Hopkins University School, The Johns Hopkins University, Baltimore, MD, United States, ²F.M.Kirby Research Center for Functional Brain Imaging, Kennedy Krieger Institute, Baltimore, MD, United States, ³Cellular Imaging Section, Institute for Cell Engineering, The Johns Hopkins University, Baltimore, MD, United States

CEST imaging is an emerging new technology with great promise for molecular imaging applications. Unfortunately, current CEST imaging schemes produce images with a large signal losses due to interference of conventional magnetization transfer and direct water saturation effects, especially for in vivo. We have developed a new method in which cosine-modulated dual-band pulses are inserted into the saturation module. These pulses shift the water-dip in Z-spectra, providing a reduction in conventional magnetization transfer effects. We have collected images on both DIACEST phantoms and 9L tumors in vivo, with the CNR of the resulting images improved over previous methods.

Olga Ivchenko^1,2, Petra Imhof², Moritz Zaiss¹, and Peter Bachert^1
1Medical Physics in Radiology, German Cancer Research Center (DKFZ), Heidelberg, Germany, ²Molecular Biophysics, Interdisciplinary Center for Scientific Computing (IWR), Heidelberg, Germany

The chemical exchange rate was determined for creatine – NH2 group (kex =102.9 s-1) at the pH=6.5 using quantum mechanics/molecular mechanics simulations (QM/MM) in combination with umbrella sampling simulations. This value is in agreement with the value from CEST experiment. The chemical exchange rate was determining using Arrhenius equation: kex =Aexp(-Ea/KbT), where A is pre-exponential factor which determines the frequency of collusion ( calculated from classical MD simulations), and Ea is an activation energy characterizing the height of a barriers which protons need to overcome in order to transfer from solute to water and vice versa (determined from QM/MM simulations).

Feliks Kogan¹, Anup Singh¹, Kejia Cai¹, Mohammad Haris¹, Hari Hariharan¹, and Ravinder Reddy^1
1Center for Magnetic Resonance and Optical Imaging, University of Pennsylvania, Philadelphia, PA, United States

Chemical Exchange Saturation Transfer (CEST) has become a popular method for measurement of metabolites with exchangeable protons. Current CEST methods are plagued by direct saturation of water protons which leads to decreased SNR and CEST contrast. In this work, we demonstrate the feasibility of using an off-resonance spin-lock (SL) pulse for generating the exchange mediated contrast. We showed in phantom experiments that this new method can decrease direct water saturation and increase chemical exchange contrast. We also demonstrated the feasibility of using the SL method in vivo in human patellar cartilage and showed that the SL method was 15% more sensitive than conventional CEST.

Sheng-Min Huang¹, Shang-Yueh Tsai², Teng-Yi Huang³, Yi-Chun Wu⁴, and Fu-Nien Wang^1
1Department of Biomedical Engineering and Environmental Sciences, National Tsing Hua University, Hsinchu, Taiwan, ²Graduate Institute of Applied Physics, National Chengchi University, Taipei, Taiwan, ³Department of Electrical Engineering, National Taiwan University of Science and Technology, Taipei, Taiwan, ⁴Molecular Imaging Center, Chang Gung Memorial Hospital, Taoyuan, Taiwan

The effect of image acquisition RF pulses on CEST imaging was investigated in this study. RARE and spin-echo EPI were utilized on a phantom experiment. The spillover effect of multiple 180 refocusing RFs was confirmed on the observed Z-spectra. The asymmetry of magnetic transfer ratio was thus reduced while using RARE for imaging acquisition. It is suggested that care must be taken when using RF pulses train for image readout.

Moritz Zaiss¹, Olga Ivchenko¹, and Peter Bachert^1
1Dpt. of Medical Physics in Radiology, German Cancer Research Center (DKFZ), Heidelberg, Baden-Württemberg, Germany

Chemical exchange saturation transfer (CEST) effect of endogenous CEST agents with small frequency offset from water resonance like amide protons (APT) are always diluted by direct saturation of water protons (spillover effect). An analytical spillover correction is presented which allows reconstructing the unperturbed pulsed-proton transfer rate for different flip angles and B₁ using only label and reference scan. Additionally, the artificial pulsed-proton transfer rate is shown to be correlated with the analytical cw proton transfer rate. Therefore, the spillover correction provides analytic quantification of pulsed saturation transfer experiments which are feasible in clinical scanners.

Qi Liu^1,2, Zhaoyang Fan², Wafa Tawackoli², Gadi Pelled², Dan Gazit², Yutaka Natsuaki³, and Debiao Li^2,4
1Northwestern University, Chicago, IL, United States, ²Cedars-Sinai Medical Center, Los Angeles, CA, United States, ³Siemens Healthcare, Los Angeles, CA,⁴University of California, Los Angeles, CA, United States

Balanced steady-state-free-precession (bSSFP) allows rapid imaging with high SNR efficiency. However, it is sensitive to scanner frequency drift, especially when long imaging time is required such as in CEST imaging. We developed an alternating-offset-saturation bSSFP (AOS-bSSFP) technique for more accurate CEST detection, as demonstrated by excellent agreement with TSE CEST in MTR _asym plot both in phantom and in vivo scans.

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

A rescaling property for simulated direct water saturation maps for pulsed RF irradiation at different frequency offsets is introduced. It allows prediction of regions of low spillover effect for a specific pulse train for all offsets of interests by simply rescaling one simulated spillover map. Additionally, for each simulated pulse power and duration the same map shows to provide the whole symmetric z-spectrum. This makes the search of low spillover regions quick and easy which is interesting for optimization of pulsed magnetization transfer and CEST experiments where small chemical shifts relative to the water protons lead to strong signal dilution by spillover effect.

Jing Yuan¹, Dong Liang², Feng Zhao¹, Yujia Li¹, Yi-Xiang J Wang¹, and Leslie Ying^3
1Imaging and Interventional Radiology, The Chinese University of Hong Kong, Shatin, NT, Hong Kong, ²Paul C. Lauterbur Research Center for Biomedical Imaging, Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen, Guangdong, China, ³Department of Electrical Engineering and Computer Science, University of Wisconsin, Milwaukee, WI, United States

T1ñ relaxation has potentials for investigating low frequency motional physiological processes but suffers from long scan time, susceptibility to motion and high SAR. We present the use of k-t ISD, a recently proposed compressed-sensing method for T1ñ mapping. The T1ñ maps of rat brains derived with and without CS were compared. The results show that T1ñ map can be accurately estimated from the highly compressed data with the reduction factor up to 10 in brain ROIs and 6 in muscle ROIs without significant differences. This technique is promising for accurate T1ñ quantification with remarkably reduced scan time and SAR.

Tobias Wech¹, Daniel Stäb¹, Andre Fischer¹, Dietbert Hahn¹, and Herbert Köstler^1
1Institute of Radiology, University of Würzburg, Würzburg, Germany

Compressed Sensing reconstructions exploiting spatio-temporal sparsity have successfully been applied to accelerate dynamic MRI. However, the non-linear and non-stationary CS algorithms prohibit the straightforward evaluation of the temporal resolution through a single global temporal point spread function. In this work a pixelwise perturbation strategy was utilized to assess local temporal point spread functions for every image pixel. The method therefore allows an appropriate assessment of the temporal resolution and can thus improve the choice of sampling patterns and algorithm settings.

James A. Rioux^1,2, Steven D. Beyea^2,3, and Chris V. Bowen^2,3
1Department of Physics, Dalhousie University, Halifax, Nova Scotia, Canada, ²Institute for Biodiagnostics (Atlantic), National Research Council, Halifax, Nova Scotia, Canada, ³Departments of Physics, Radiology and Biomedical Engineering, Dalhousie University, Halifax, Nova Scotia, Canada

TurboSPI can be used to acquire a time series of images suitable for high temporal resolution relaxometry, but must be significantly accelerated to permit applications such as cellular imaging in vivo. The acquisition of a matched Fast Spin Echo image provides prior information to further sparsify the reconstructed image. We compare two approaches to incorporating this prior information, and show that a modified-CS reconstruction using the known support of the FSE template allows acceleration factors of up to 30 while retaining high image quality throughout the time series.

Cagdas Bilen¹, Ricardo Otazo², Daniel K Sodickson², Ivan Selesnick¹, and Yao Wang^1
1Department of Electrical Engineering, Polytechnic Institute of NYU, Brooklyn, NY, United States, ²Bernard and Irene Schwartz Center for Biomedical Imaging, NYU School of Medicine, New York, NY, United States

Video coding techniques such as motion compensation has been proposed to exploit temporal redundancy and improve compressed sensing reconstructions of undersampled dynamic MRI data. Many of these methods require reference frames and/or fully sampled low pass k-space data which limits the acceleration factor. We propose a regularization framework with motion compensating prior that adaptively estimates the motion field during the reconstruction iterations with no need for reference frames or fully sampled k-space data.

Ricardo Otazo¹, Riccardo Lattanzi¹, and Daniel K Sodickson^1
1Bernard and Irene Schwartz Center for Biomedical Imaging, NYU School of Medicine, New York, NY, United States

The limits of acceleration for combinations of compressed sensing and parallel imaging remain uncertain. In this work, we investigate the performance of the combined reconstruction with respect to the number of coils for truly-sparse and compressible MR images. A complete basis set of electromagnetic fields is employed as a hypothetical optimal coil array, which achieves the best possible SNR for parallel imaging reconstructions. We demonstrate that the minimum number of required k-space samples is bounded by the number of sparse coefficients, which removes the oversampling factor of 3-5 for compressed sensing alone and approximates the theoretical bounds of L0-norm minimization.

Anja Lutz¹, Sebastian Kozerke², Jan Paul¹, Michael Schmucker¹, Axel Bornstedt¹, G Ulrich Nienhaus³, Wolfgang Rottbauer¹, and Volker Rasche^1
1University Hospital of Ulm, Ulm, BW, Germany, ²ETH Zürich, Institute for Biomedical Engineering, Zürich, Switzerland, ³Karlsruhe Institute of Technology, Karlsruhe, BW, Germany

Quantification of myocardial mechanics by tissue phase mapping (TPM) is supposed to provide an improved understanding of cardiac motion and to enable a detailed assessment of myocardial diseases such as cardiac asynchrony and dilated cardiomyopathy. A major limitation of TPM in clinical routine is the long acquisition time. In this study different reconstruction algorithms (k-t BLAST, k-t SENSE, k-t PCA and k-t PCA/SENSE) are investigated regarding their applicability to accelerated TPM data. For high acceleration factors, the use of sensitivity maps and principal component analysis improves the quality of velocity data.

Haonan Wang¹, Neal Kepler Bangerter^1,2, Ganesh Adluru², Matthias Schabel³, Glen R. Morrell², and Edward V.R. DiBella^2
1Electrical & Computer Engineering, Brigham Young University, Provo, UT, United States, ²Department of Radiology, University of Utah, Salt Lake City, UT, United States, ³Ohio State University

Dynamic contrast-enhanced magnetic resonance imaging (DCE MRI) of breast tumors provides a promising method for the evaluation of vessel permeability in the tumor area. Recently, a number of constrained reconstruction algorithms using Total Variation (TV) and Low Rank (LR) have been proposed to mitigate the tradeoff between spatial and temporal resolution in DCE MRI. In this work, we compare two of these acceleration methods on DCE MRI of the breast. We demonstrate that high acceleration rates (R=13) are potentially feasible in DCE-MRI of the breast using both the TV and LR constrained reconstruction algorithms.

Claudio Santelli^1,2, Sebastian Kozerke^1,2, and Tobias Schaeffter^2
1Institute for Biomedical Engineering, University and ETH Zurich, Zurich, Switzerland, ²Division of Biomedical Engineering and Imaging Sciences, King's College London, London, United Kingdom

Among the various scan acceleration techniques available to Phase-Contrast(PC)-MRI, Compressed sensing (CS)has recently been demonstrated. In this work, it It is hypothesized that reconstruction accuracy with respect to object phase depends on inflow contrast. Using 2D and 3D in-vivo flow data acquired in the aortic arch it is demonstrated that reconstruction error varies significantly on the amplitude contrast.

Yijing Wu¹, Kevin M Johnson¹, Steven R Kecskemeti², Patrick A Turski³, and Chuck A Mistretta^4
1Medical Physics, University of Wisconsin, Madison, WI, United States, ²University of Wisconsin, ³Radiology, University of Wisconsin, Madison, WI, United States, ⁴Medical Physics and Radiology, University of Wisconsin, Madison, WI, United States

Hybrid HYPR-MRA utilizes a separately acquired high spatial resolution static MRA as a spatial constraint to reconstruct highly undersampled dynamic MRA to achieve high temporal and spatial resolution simultenously. However, for extremely high undersampling factor (>100) or/with reduced sparsity, severe undersampling artifacts not only contaminate the image, but also cause signal "crosstalk" from distant vessels, which appears as early enhanced veins or prolonged enhanced arteries. In this project, we integrate Parallel imaging and CS to eliminate undersampling artifacts before HYPR algorithm, such that the waveform fidelity of small vessels can be improved, at the same time, CS related artifacts and SNR can be improved by Hybrid HYPR technique.

Hersh Chandarana¹, Li Feng¹, Tobias Kai Block¹, Andrew B Rosenkrantz¹, Ruth P Lim¹, Dewey Chu¹, Daniel K Sodickson¹, and Ricardo Otazo^1
1Radiology, NYU Langone Medical Center, New York, NY, United States

Dynamic post-contrast liver MR examination is performed with Cartesian k-space sampling in a breath-hold (BH). However, this results in non-diagnostic images in patients who cannot adequately breath-hold. Purpose of this study was to compare image quality of free-breathing radial (interleaved ‘angle-bisection’ and golden angle) acquisition schemes reconstructed with either view sharing (KWIC) or joint compressed-sensing and parallel imaging reconstruction (CS-PI), to conventional BH exam. Our results demonstrate that free-breathing radial golden angle trajectory continuous acquisition scheme with CS-PI reconstruction has image quality comparable to a breath-hold exam and significantly better than interleaved radial acquisition with KWIC or CS-PI reconstruction.

Angshul Majumdar¹, Rabab K Ward², and Tyseer Aboulnasr^3
1Electrical and Computer Engineering, University of British Columbia, Vancouver, British Columbia, Canada, ²Electrical and Computer Engineering, University of British Columbia, Vancouver, BC, Canada, ³Electrical and Computer Engineering, University of British Columbia

A method for reconstructing dynamic MRI sequence, acquired via partial parallel imaging is proposed here. The novelty of the method is that it does not require explicit or implicit knowledge of the sensitivity maps. The method is compared against k-t BLAST and k-t GRAPPA - both of which the sensitivity profile in one form or the other. Our method is as good as k-t BLAST and better than k-t GRAPPA.

Joshua D. Trzasko¹, Yunhong Shu², Armando Manduca¹, John Huston III², and Matt A Bernstein^2
1Physiology and Biomedical Engineering, Mayo Clinic, Rochester, MN, United States, ²Department of Radiology, Mayo Clinic, Rochester, MN, United States

In non-Cartesian image reconstruction, equality data constraints cannot be explicitly enforced, which limits the applicability of popular methods like projection-onto-convex-sets (POCS) to this area. In this work, we demonstrate that they can be implicitly enforced via a specific affine projection, and discuss numerical methods for efficiently imposing them. We use this construction to develop an efficient Compressive Sensing reconstruction based on block-wise redundant sparsity constraints, which results in strong reconstruction performance. We demonstrate the proposed reconstruction strategy for CE-MRA acquired using the 3D SWIRLS trajectory.

Yoon-Chul Kim¹, Nassos Katsamanis¹, Michael Proctor¹, Shrikanth Narayanan¹, and Krishna S. Nayak^1
1Electrical Engineering, University of Southern California, Los Angeles, CA, United States

In the context of vocal tract imaging during fluent speech, golden-ratio (GR) spiral imaging has been shown to provide improved depiction of articulators over conventional spiral imaging. However, one drawback of GR imaging is incompatibility with MRI-gradient noise cancellation methods because there is no periodicity in the gradient waveforms. We apply GR spiral view order with fixed number of interleaves (i.e. pseudo GR spiral) to promote periodicity. The proposed method is well-suited to parallel imaging for acceleration thus leading to improved image quality for selection of a small temporal window and also provides adequate sound quality (and/or comparable speech alignment accuracy) compared to the conventional method.

Andrew David Scott¹, Redha Boubertakh¹, Malcolm Birch¹, and Marc Eric Miquel^1
1Clinical Physics, Barts and the London NHS Trust, London, Greater London, United Kingdom

Previous real-time MR studies of soft palate motion have had limited frame-rates or used complex spiral techniques. We implement and compare high frame-rate (9-20fps) 1.5T bSSFP and 3.0T SSFP sequences. Comparisons were made in healthy subjects and images were synchronised with simultaneously acquired audio recordings. SNR was higher in 3.0T acquisitions at rest, but lower during speech compared to 1.5T. 3.0T images were reliably of diagnostic quality. 1.5T images were mostly of the highest quality but frequently undiagnostic. Comparison between sequences suggested that for the assessment of soft-palate motion, high frame-rates (~20fps) are desirable at the expense of some spatial-resolution.

Tobias Hahn¹, Sebastian Kozerke¹, Ruben Pellicer Guridi^1,2, Werner Schwizer³, Michael Fried³, Peter Boesiger¹, and Andreas Steingoetter^1
1Institute for Biomedical Engineering, University and ETH Zurich, Zurich, Switzerland, ²Department of Biophysics and Bioengineering, University of Barcelona, Barcelona, Spain, ³Dep. of Internal Medicine, Division of Gastroenterology and Hepatology, Zurich, Switzerland

This study presents the simultaneous in vivo tracking of three identical fluorine labeled capsules using an improved 3D Golden Angle based sampling scheme. Liquid fluorine marker was Perfluoro-15-crown-5-ether and capsule filling was 65µl each. In vivo SNR for different reconstruction window sizes is given. Furthermore, the effect of constrained reconstruction onto the tracking reliability is studied, by multiplication of reconstructed 3D images covering a large temporal range with dynamic 19F images. SNR comparable to other tracking methods is found and good tracking reliability from a temporal resolution of 220 ms.

Samantha J Holdsworth¹, Anh T Van¹, Stefan Skare², and Roland Bammer^1
1'Center for Quantitative Neuroimaging, Department of Radiology, Stanford University, Palo Alto, CA, United States, ²Clinical Neuroscience, Karolinska Institute, Stockholm, Sweden

Readout-Segmented EPI (RS-EPI) has been proposed in a number of studies as a variant of EPI to reduce distortion in diffusion-weighted (DW) neuroimaging. With the increasing number of coils coming available with advanced phased-array technology that increases the capacity of EPI to achieve higher acceleration factors, and with advanced distortion-correction methodology, we present preliminary data that shows that EPI may be more useful for acquisition of DWI images in clinically-acceptable scan times. Here we compare EPI and RS-EPI data acquired with the highest acceptable GRAPPA-acceleration factor that we have been using with our 32-channel coil.

Christian Horn^1,2, Vladimir Juras^1,3, Stefan Ropele⁴, and Andreas Berg^1,2
1MR-Center of Excellence, Medical University of Vienna, Vienna, Vienna, Austria, ²Center for Medical Physics and Biomedical Engineering, Medical University of Vienna, Vienna, Vienna, Austria, ³Department of Radiology, Vienna General Hospital, Medical University of Vienna, Vienna, Vienna, Austria,⁴Department of Neurology, Medical University of Graz, Graz, Steiermark, Austria

The combination of special, radial pulse sequences designed for the detection of ultra-short echo times (UTE, TEmin = 0.07 ms) with the clinically more established technique of magnetization transfer contrast promises a new image contrast in semi-solid tissue components or implants with very short T2 times (< 1 ms). A corresponding pulse sequence was evaluated on a custom designed MR microscopy gradient insert on a 7T human scanner. BSA phantoms, with different BSA to water concentrations, and biological samples were used to demonstrate the practical applicability of the sequence ex vivo.

Jing-Tzyh Alan Chiang¹, Michael Carl², Jiang Du¹, and Graeme Bydder^1
1Radiology, University of California San Diego, San Diego, CA, United States, ²GE Healthcare

We investigate the signal amplitude relationships due to T2 decay during radial k-space readout in ultrashort TE sequences, via numerical simulations on 2D rings of varying inner and outer diameters. Decreasing inner and increasing outer diameters lead to increased signal amplitudes at any given T2 value, which demonstrate important dependencies of signal amplitude on object shape and size. These results on 2D rings also serve as numerical models for analyzing T2 dependent signal and contrast in quantitative UTE techniques that utilize axial imaging of tubular bones, such as in cortical bone water quantification.

Masami Yoneyama¹, Masanobu Nakamura¹, Tomoyuki Okuaki¹, Takashi Tabuchi¹, Atsushi Takemura², Tetsuo Ogino², Makoto Obara², Thomas Kwee³, and Taro Takahara^4
1Yaesu Clinic, Tokyo, Japan, ²Philips Electronics Japan, Tokyo, Japan, ³University Medical Center Utrecht, Utrecht, Netherlands, ⁴Tokai University School of Engineering, Kanagawa, Japan

Contrast-enhanced 3D-T1-weighted imaging based on GRE is widely used for detecting small brain metastases, but since contrast materials remain in both blood and the tumor parenchyma and thus increase the signal intensity of both regions, it is often challenging to distinguish brain tumors from blood. To overcome this problem, "black-blood" version of T1 weighted images based on TSE with/without MSDE are used recently, but these methods are difficult to "perfect" suppression of whole-vessel signals, and therefore the differentiation of residual blood vessel and small brain metastasis is occasionally difficult. In this study, we propose a new scheme of fast, volumetric, high-resolution, bright- and black-blood imaging simultaneous acquisition (VISIBLE). This optimal sequence can be used for 3D volumetric T1 weighted bright- and black-blood imaging, and that is promising for detecting small brain metastases, by the differentiation improvement of blood vessel and small brain metastasis.

Manojkumar Saranathan¹, Michael Zeineh¹, Geoffrey A. Kerchner², Mohammad Mehdi Khalighi³, Marcus T. Alley¹, and Brian K. Rutt^1
1Radiology, Stanford University, Stanford, CA, United States, ²Neurology and Neurological Sciences, Stanford University, Stanford, CA, United States,³Global Applied Science Labatory, GE Healthcare, Menlo Park, CA, United States

Fast Spin Echo (FSE) imaging is increasingly used for fast T2 weighted imaging applications. However, at 7T, the sequence is heavily limited by SAR considerations, severely reducing its time efficiency. Flip angle modulation schemes like SPACE and XETA [1-2], have been proposed for 3D FSE that enable the use of longer echo-train-lengths and help lower SAR due to reduced refocusing flip angles. We optimized the 3D XETA refocusing flip angle train for T2 weighted brain imaging at 7T based on SAR, signal intensity, contrast and point spread function (PSF) considerations, using T1 and T2 values of white/grey matter at 7T. Additionally, we explored the use of composite excitation pulses to mitigate signal loss from B1 inhomogeneity effects. Whole brain 3D T2 imaging was performed on patients using these modifications.

Ann-Kathrin Homagk¹, Alexander Radbruch², and Wolfhard Semmler^1
1Dept. of Medical Physics in Radiology, German Cancer Research Center (DKFZ), Heidelberg, Germany, ²Dept. of Neuroradiology, University Heidelberg, Heidelberg, Germany

In this work, we investigate the suitability of different ASL techniques for high resolution dynamic angiography imaging of the cerebral vasculature at 7 Tesla. As the basic STAR sequence is prone to magnetization transfer effects, an additional inversion pulse was inserted prior to the acquisition of the control image. Furthermore, a PICORE, a FAIR and a FAIRER sequence were implemented. The cerebral vasculature of three volunteers was imaged with an in-plane resolution of 0.57x0.57mm² using all five sequences. The acquisition time for the magnetization transfer compensated techniques was increased by up to 20%; however the visibility of peripheral vessels was improved.

Kieran O'Brien^1,2, Jean Delacoste¹, Jose Marques¹, Tobias Kober³, Francois Lazeyras², Rolf Gruetter¹, and Gunnar Krueger^3
1LIFMET-CIBM, Ecole Polytechnique Fédéral de Lausanne, Lausanne, Switzerland, ²Department of Radiology-CIBM, Université de Genève, Geneva, Switzerland, ³Advanced Clinical Imaging Technology, Siemens Suisse, Switzerland

In whole brain acquisitions, such as MP2RAGE, large B0 offsets or regions of low B1 can suffer from poor inversion. Optimisation of adiabatic pulses to work at lower B1, allows for better inversion profiles extending the coverage of whole brain acquisitions.

Rajiv G Menon¹, Bradley P Sutton², Donald B Twieg³, and Timothy J Carroll^1
1Department of Radiology, Feinberg School of Medicine, Northwestern University, Chicago, IL, United States, ²Bioengineering Department, University of Illinois at Urbana-Champaign, Urbana, IL, United States, ³Biomedical Engineering Department, University of Alabama at Birmingham, Birmingham, AL, United States

We propose a novel technique for the measurement of oxygen extraction fraction (OEF) using Parameter Assessment by Retrieval from Signal Encoding (PARSE) techniques. OEF is a key indicator of cerebral perfusion and gives crucial information regarding the stage of vascular compromise in patients with neurovascular disease. PARSE is an efficient, multi-parameter estimation technique, that uses a more accurate signal model to simultaneously measure M0, R2* and local frequency. By direct measurement of local frequency changes due to OEF related susceptibility changes using PARSE, we determine the feasibility of robustly estimating OEF in pathological situations.

Lea E Marais¹, Jean-François Mangin^2,3, Cyril Poupon^2,3, Urielle Thoprakarn^2,3, Vincent Perlbarg⁴, Gareth J Barker⁵, Xavier Golay⁶, Joseph V Hajnal⁷, Adam J Schwarz ⁸, and Derek Hill^1
1IXICO Ltd., London, United Kingdom, ²I²BM NeuroSpin, CEA, Gif-sur-Yvette, France, ³CATI, Gif-sur-Yvette, France, ⁴UMR-S 678, Laboratoire d'Imagerie Fonctionnelle, Inserm and UPMC Univ Paris06, Paris, France, ⁵Centre for Neuroimaging Sciences, King's College London, Institute of Psychiatry, London, United Kingdom, ⁶MR Neurophysics and Translational Neuroscience, UCL Institute of Neurology, National Hospital for Neurology & Neurosurgery, London, United Kingdom, ⁷Robert Steiner MRI Unit,Imaging Sciences Department, MRC Clinical Sciences, Hammersmith Hospital, Imperial College London, London, United Kingdom, ⁸Translational Medicine, Eli Lilly and Company, Indianapolis, United States

Diffusion Tensor Imaging (DTI) and resting state functional Magnetic Resonance Imaging (rs-fMRI) biomarkers correlate with Alzheimer’s Disease stage and provide additional information on the efficacy of new treatments. Multi-centre clinical studies are essential for rapid enrolment of a sufficient number of patients. We deployed ADNI-style fMRI and DTI sequences on Philips and Siemens 3T scanners and evaluated their repeatability and reproducibility on healthy volunteers. The variability across scanners was comparable to the within-scanner variability. These sequences are thus suitable for the acquisition of comparable data in multi-site, multi-vendor clinical trials.

Na Zhang^1,2, Guanghua Song^1,2, Weiqi Liao^1,2, Weijie Tao^1,2, Leslie Ying³, and Dong Liang^1,2
1Paul C. Lauterbur Research Centre for Biomedical Imaging, Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen, Guangdong, China, ²Key Laboratory of Health Informatics, Chinese Academy of Sciences, Shenzhen, Guangdong, China, ³Department of Electrical Engineering and Computer Science, University of Wisconsin-Milwaukee, Milwaukee, WI, United States

The emerging CS-based methods have shown promising performance for accelerating DCE-MRI. In this work, the potential of k-t Iterative Support Detection (k-t ISD, a recently proposed dynamic imaging method based on CS with partial known support theory) in accelerating DCE-MRI is investigated and compared to the sliding window (SW) method. The superiority of this method to conventional CS methods has been demonstrated on cardiac cine imaging. The reconstruction results and statistical analysis indicate that k-t ISD can faithfully reconstruct the uptake curves and improve temporal resolution of DCE-MRI without compromising the spatial resolution, when a high net reduction factor is used.

L. Keith¹, M. Rahimi², J. Holmes³, K. Wang³, J. Brittain³, and F. Korosec^1,4
1Medical Physics, University of Wisconsin-Madison, Madison, WI, United States, ²Biomedical Engineering, University of Wisconsin - Madison, Madison, WI, United States, ³Global Applied Science Laboratory, GE Healthcare, ⁴Radiology, University of Wisconsin - Madison, Madison, WI, United States

In this work, an experimental setup has been designed in which a computer-controlled motion stage is used to translate an object through the imaging FOV during data acquisition in an attempt to mimic a bolus of contrast material traveling through vasculature. Since the motion is computer controlled and the dynamic elements of the phantom are well-defined with sharp edges, the input to the HYPR process is well-known. We conduct two experiments: Experiment 1 investigates the temporal and spatial fidelity of HYPR processing; Experiment 2 compares the fidelity of HYPR against that of current clinical MRA techniques.

Michelle Yan¹, Jeff Johnson¹, Xiao Chen², Li Pan², Ti-Chiun Chang¹, Yunqiang Chen¹, and Tong Fang^1
1Siemens Corporate Research, Princeton, NJ, United States, ²Center for Applied Medical Imaging, Siemens Corporate Research, Baltimore, MD, United States

We propose to incorporate human perception model into assessing the quality of compressed sensing MR reconstruction algorithms. More specifically, we seek to measure the perceptual changes or degradation in reconstructed MR images using the signed just noticeable difference (JND) analysis, providing an overall “quality” color map or score for each MR image under consideration. Red color indicates positive (added) changes; blue color indicates negative (lost) changes. The lighter a JND map in color, the better a reconstruction method performs. The result has demonstrated potential value of JND maps in assessing image quality and comparing the overall performance of MR reconstruction methods.

Kyunghyun Sung¹, Anderson N Nnewihe^1,2, Bruce L Daniel¹, and Brian A Hargreaves^1
1Radiology, Stanford University, Stanford, CA, United States, ²Bioengineering, Stanford University, Stanford, CA, United States

Compressed sensing (CS) is a technique that allows accurate reconstruction of images from a reduced set of acquired data. Here, we describe two wavelet subband decomposition methods for High-frequency Subband CS, which enable to employ selective application of different undersampling patterns and reconstructions in different k-space regions. Two wavelet decomposition methods are evaluated in high-resolution T1- and T2-weighted 3D breast imaging with extremely high acceleration factors.

Julia V Velikina¹, and Alexey A Samsonov^2
1Medical Physics, University of Wisconsin - Madison, Madison, Wisconsin, United States, ²Radiology, University of Wisconsin - Madison

We propose a novel model-driven compressed sensing approach for T1 relaxometry. The proposed algorithm alternates signal estimation with adaptive update of sparsifying transform based both on the analytical signal model and current signal estimate. The proposed algorithm can also be used in other quantitative MRI applications.

Guillaume Madelin¹, Gregory Chang¹, Alexej Jerschow², Ricardo Otazo¹, and Ravinder R Regatte^1
1Radiology Department, New York University Medical Center, New York, NY, United States, ²Chemistry Department, New York University, New York, NY, United States

Quantitative sodium MRI is highly specific to the glycosaminoglycan content in cartilage and could be used to assess the biochemical degradation of cartilage in early stages of osteoarthritis. However, due to low sodium NMR sensitivity and its low concentration, sodium images need long acquisition times (15 to 25 min, respectively without and with fluid suppression) at 7T and are typically of low resolution. In this preliminary study, we show that compressed sensing can be applied to reconstruct undersampled sodium images and reduce the acquisition time by a factor of 2 at 7T without losing sodium quantification accuracy.

Sebastian Weingärtner¹, Friedrich Wetterling¹, and Lothar R. Schad^1
1Computer Assisted Clinical Medicine, University Medical Center Mannheim, University of Heidelberg, Mannheim, Germany

The purpose of this study was to develop a novel undersampling scheme, for the case that acquisition does not take place in the k-space. To achieve this, an acquisition scheme that parallels a binary search algorithm was developed. By scanning successively the sum of several coefficients in a sparse image transformation domain high undersampling factors can be reached. The algorithm was evaluated and showed only minor differences for undersampling with up to 4% of the data. These results can be achieved by an acquisition in a sparse-domain and do not need compressed sensing post-processing.

Abolfazl Mehranian¹, Hamidreza Saligheh Rad^1,2, Mohammadreza Ay^1,2, and Arman Rahmim^3
1Medical Physics and Biomedical Engineering Department, Tehran University of Medical Sciences of Medical Sciences, Tehran,Iran, Tehran, Tehran, Iran,²Research Center for Science and Technology in Medcine, Imam Hospital, Tehran, Tehran, Iran, ³Department of Radiology, School of Medicine, Johns Hopkins University, Baltimore, Maryland, United States

The compressive sensing (CS) of spirally encoded MR acquisitions makes it possible to significantly reduce the scanning time in 3D MR imaging techniques. In this work, we studied an efficient primal-dual algorithm for 3D total variation (TV) and Huber-based compressed MR image reconstruction. We tailored this algorithm for TV and Huber regularizations in 3D and made use of a stack of variable-density spiral trajectories for 80% k-space undersampling. In a volumetric cardiac dataset, it was demonstrated that the derived algorithm objectively outperforms several state-of-the-art algorithms and thus can have promising clinical implications in fast MR imaging.

Hua Wang¹, Yeyang Yu¹, Bing Keong Li¹, Adnan Trakic¹, Mingjian Hong², Feng Liu¹, and Stuart Crozier^1
1The University of Queensland, Brisbane, QLD, Australia, ²ChongQing University, ChongQing, China

In this work, we presented a method to optimise the k-space sampling scheme for CS-MRI. The problem was simplified by treating the variable density functions in parts, and optimising the weighting factors with GA. A 2D brain and a 3D apple MRI image reconstruction illustrates an improved imaging quality with this method.

Gabriel Rilling¹, Yuehui Tao², Mike E. Davies¹, and Ian Marshall^2
1School of Engineering, University of Edinburgh, Edinburgh, Scotland, United Kingdom, ²Medical Physics, University of Edinburgh

A novel model-based undersampling and reconstruction framework is proposed for phase contrast carotid blood velocity mapping. It is based on splitting the space-frequency support of the signal into two elementary blocks and use a combination of two sampling lattices adapted to each block. This multi-lattice sampling approach is intermediate between lattice and random sampling, allowing more flexibility and acceleration than the former and better noise robustness than the latter. Combined with the "keyhole" technique, it allows up to 12X acceleration with limited distortion. Simulation and in vivo undersampling experiments at 12X and 8X acceleration validate the proposed method.

Jian-Xiong Wang^1
1Applied Science Laboratory, GE Healthcare, London, ON, Canada

This work demonstrated that true 3D, or Cubic, Compressed Sensing reconstruction method not only reduced the computation time by about a factor of two, but also greatly increased image quality for Sparse MRI.

Alicia W Yang^1,2, Li Feng^1,2, Jian Xu^3,4, Ivan Selesnick⁴, Daniel K Sodickson¹, and Ricardo Otazo^1
1Department of Radiology, New York University School of Medicine, New York, NY, United States, ²Sackler Institute for Biomedical Sciences, New York University, New York, NY, United States, ³Siemens Medical Solution USA Inc, New York, United States, ⁴Polytechnic Institute of New York University, New York, United States

An adaptive decreasing thresholding method is developed to take into consideration the structure of overcomplete transforms by (1) using local thresholds that adapt to the signal power in each band and (2) decreasing the threshold for each step of the iterative reconstruction algorithm. The performance of this method in dual-tree wavelet transforms and curvelets was tested on compressed sensing reconstructions of retrospectively undersampled 3D coronary MRA and brain image datasets, and compared to that of standard Haar wavelet transforms.

Yoon-Chul Kim¹, R. Marc Lebel¹, Michael C.K. Khoo², and Krishna S. Nayak^1
1Electrical Engineering, University of Southern California, Los Angeles, CA, United States, ²Biomedical Engineering, University of Southern California, Los Angeles, CA, United States

Imaging of upper airway dynamics can give new insights into the mechanisms of airway obstruction and provide data for its modeling. Conventional MRI is not fast enough to capture the entire airway volume and its dynamics. We demonstrate dynamic 3D imaging of the airway during inspiratory loading using 3D Cartesian imaging with golden-angle temporal view order and L1-SPIRiT reconstruction, to achieve 1.6 mm isotropic spatial and sub-second temporal resolution. The end result is a first-ever visualization of airway collapse using real-time 3D MRI.

L.C. Bell¹, K.M. Johnson², S.B. Fain¹, S.J. Kruger¹, and S.K. Nagle^1,2
1Medical Physics, University of Wisconsin-Madison, Madison, Wisconsin, United States, ²Radiology, University of Wisconsin-Madison, Madison, Wisconsin, United States

Various techniques currently in place for T1 mapping of the lung parenchyma are limited by the low signal in the lungs. Recent developments have shown that sequences utilizing an ultra-short echo time are capable of acquiring MR signal in small animal lung parenchyma. We propose to use a 3D Radial UTE acquisition in conjunction with DESPOT1 to produce T1 maps of human lung parenchyma with sufficient signal while still maintaining high isotropic spatial resolution for functional and structural analysis.

Bing Wu¹, Wei Li², and Chunlei Liu^2
1GE heathcare, Beijing, China, ²Brain Imaging and analysis center, Duke University

Susceptibility imaging with SPGR is a considerably long imaging process which limits its clinical feasibility. Acceleration with parallel imaging, however, degrades the SNR in the resulting susceptibility map. In this work, additional data acquisitions are made in the otherwise wasted empty echo space in front of the readout. The SNR improvement gained through multi-echo averaging has been shown to completely offset the SNR loss incurred at an acceleration factor of 4, showing visibly even better SNR than fully sampled dataset at a single TE.

Bing Wu¹, Wei Li², Nankuei Chen², and Chunlei Liu^2
1GE heathcare, Beijing, China, ²Brain Imaging and analysis center, Duke University

Susceptibility imaging with standard SPGR is a considerably long imaging process which limits its clinical feasibility. Spiral acquisition has been proposed as an alternative to achieve rapid susceptibility imaging. However, spiral trajectory places demanding requirements on the gradient system and is not usually available on commercial scanners. In addition, spiral trajectory is very sensitive to susceptibility distortions. In this work, a multi-echo multi-shot EPI sequence is implemented and used for susceptibility imaging. The result shows that, given the same acquisition time, EPI leads to a susceptibility map with less distortion compared to that of spiral acquisition.

Jennifer E Dixon¹, Ashley I Simpson², Niraj Mistry², Nikos Evangelou², and Peter G Morris^1
1Sir Peter Mansfield Magnetic Resonance Centre, University of Nottingham, Nottingham, Nottinghamshire, United Kingdom, ²Academic Division of Clinical Neurology, University of Nottingham, Nottingham, Nottinghamshire, United Kingdom

7T T2*-weighted MRI has been shown to allow differentiation between MS lesions and microangiopathic white-matter (WM) lesions by the presence of a visible vein, and accurately predicts eventual diagnosis in patients who were otherwise unable to be diagnosed without further testing; however, the limited availability of 7T systems necessitates the translation of this technique to lower field. The aim of this work is to increase its sensitivity at 3T and 7T, and to compare the optimised sequences theoretically and experimentally in order to quantify the benefit of imaging at higher field, and to enable its use at clinical field strengths.

Wei Liu^1,2, Qun Zhao³, Haiying Tang⁴, Binquan Wang^1,2, Ping-Hong Yeh^1,2, Dominic E Nathan^1,2, John Graner¹, Hai Pan^1,2, Rachel Wolfowitz^1,2, Jamie Harper¹, Louis M French⁵, Terry R Oakes¹, and Gerard Riedy^1
1National Neuroimaging Consortium, National Intrepid Center of Excellence, Bethesda, MD, United States, ²Henry M Jackson Foundation for the Advancement of Military Medicine, Bethesda, MD, United States, ³Department of Physics and Astronomy, University of Georgia, Athens, GA, United States,⁴Uniformed Services University of the Health Sciences, Bethesda, MD, United States, ⁵Walter Reed National Military Medical Center, Bethesda, MD, United States

Two positive contrast techniques using phase gradient mapping (PGM) and susceptibility gradient mapping (SGM) were investigated as complementary approaches of susceptibility weighted imaging (SWI) for microhemorrhage detection in patients with combat related traumatic brain injury. The results demonstrated that PGM and SGM improved the CNR for the local regional of interest and might provide additional information for diagnosis.

Ziqi Sun¹, Sergey Petryakov¹, Wael F Alzawahra¹, and Jay L Zweier^1
1Davis Heart and Lung Research Institute, The Ohio State University, Columbus, Ohio, United States

T1-weighted (T1W) contrast enhancement using free radical probe as the contrast agent is applied for EPRI/MRI co-imaging of isolated rat hearts. Hyper-intense signal intensity was observed in the right-ventricle of the rat heart in the 3D T1W image, which is consistent with the greater signal intensity shown in the low resolution 3D EPRI image overlaid onto the high resolution 3D T1W image. This result indicates that high resoltion T1-weighted MRI technique is useful in accurately dipicting free radical probe distribution within tissue in EPRI/MRI co-imaging experiments.

Ronald J. Beyers¹, Nouha Salibi^1,2, Rajesh Amin³, John C Quindry⁴, and Thomas S Denney^1
1MRI Research Center, Electrical Engineering, Auburn University, Auburn, AL, United States, ²Healthcare, USA, Siemens Corporation, United States,³Harrison School of Pharmacy, Auburn University, Auburn, AL, United States, ⁴Department of Kinesiology, Auburn University, Auburn, AL, United States

Introduction: Need for cardiac MRI in rodents on clinical scanner with: 1) bright & dark blood cine, and 2) inversion recovery (IR), late gadolinium enhanced (LGE) & look-locker (LL) T1-mapping. Methods: Sequence developed ran on a Siemens 3T Verio. Included adjustable TR loops, multislice GRE, and RF preparations including: 1) DIR for dark blood, and 2) IR prep. Imaged C57BL/6 mice and Wistar rats, anesthesized by isoflurane, and ECG/respiratory tiggered. Results: Cine time was 4-6 minutes. Wrist coil provided sufficient signal. Cine provided good volumetrics and wall thickness. IR LL on a water-deprived mouse estimated myocardial T1 = 662 ms. The IR LGE made 6-slice acquisition in 13 minutes.

Henk Smit¹, A. Ruggiero¹, G.N. Doeswijk¹, M. Milanesi^2,3, G.C. Houston⁴, M.R. Bernsen¹, G.P. Krestin¹, S. Klein¹, and G. Kotek^1
1Erasmus MC, Rotterdam, Zuid Holland, Netherlands, ²Agilent Technologies UK Ltd., ³MRI Lab, Fondazione G. Monasterio-CNR, Pisa, Italy, ⁴GE Healthcare, Netherlands

This work addresses the feasibility, accuracy and the sensitivity of a novel T1 mapping method in rat myocardium at 7.0 T. With the modified Cine Inversion Recovery (mCINE-IR), the effective TR can be adjusted to allow much higher longitudinal recovery between two subsequent inversions pulses than with conventional methods. Furthermore the magnetization recovery is monitored by the CINE loop acquisition for the entire number of RR intervals within the chosen TR. The results show that mCINE-IR offers a reproducible T1 estimation of both healthy tissue and injected Gd labelled cells in the rat myocardium at 7T making longitudinal studies feasible.

Takashi Nishihara¹, Hikaru Hanada¹, Kuniharu Oka¹, Masahiro Takizawa¹, Chikako Moriwake², Hiroyuki Itagaki¹, Tetsuhiko Takahashi¹, Yosuke Hirata³, and Makoto Sasaki^4
1MRI System Division, Hitachi Medical Corporation, Kashiwa, Chiba, Japan, ²Marketing Division, Hitachi Medical Corporation, Tokyo, Japan, ³Center for Radiological Sciences, Iwate Medical University Hospital, Morioka, Iwate, Japan, ⁴Division of Ultrahigh Field MRI, Institute for Biomedical Sciences, Iwate Medical University, Morioka, Iwate, Japan

Non-gated T1W plaque imaging has been proposed for evaluating plaque characteristics. To improve image quality, we introduced flow dephasing pulses to eliminate residual blood signal and evaluated it. We determined the b value of flow dephasing pulses which can eliminate flow signal, and the ratio P of the asymmetric sampling that doesn�ft affect spatial resolution. We applied the determined P value of 8% and b value of 0.4s/mm² to volunteer imaging. The residual flow signal in ICA was eliminated and image resolution was not affected. We achieved a good contrast for plaque imaging by using non-gated radial sampling SE sequence.

Neville D Gai¹, and John A Butman^1
1Radiology & Imaging Sciences, National Institutes of Health, Bethesda, MD, United States

Alternating excitation and/or phase perturbed balanced SSFP provides highly localized variations in magnetization. The technique has been recently proposed as a means to provide positive contrast imaging. Here, the inflow effect is studied in relation to the technique. Simulations and in-vivo imaging in the brain show that blood as well as CSF pulsatility can be a confounding factor when using the sequence for positive contrast imaging. Means to reduce inflow effects need to be developed.

Lukas Pirpamer¹, Gernot Reishofer², and Stefan Ropele^1
1Department of Neurology, Medical University of Graz, Graz, Styria, Austria, ²Department of Radiology, Medical University of Graz, Graz, Styria, Austria

Non-linearities of gradient fields can cause a shift of the modulation function when SPAMM is used for image tagging. For certain application such as high resolution imaging with the microSPAMM technique, these shifts can seriously affect image reconstruction even if they are small. In this experimental work a new method for assessing these gradient induced variations of the modulation function is proposed. First phantom measurements demonstrate the feasibility and the advantage over existing methods.

Peng Lai¹, and Anja C.S Brau^1
1Global Applied Science Laboratory, GE Healthcare, Menlo Park, CA, United States

This work develops a novel source data decoupling method for computationally efficient auto-calibrating parallel imaging for arbitrary Cartesian sampling and validates its performance on 3D cardiac cine MRI with different acceleration factors and different coils. Our computation analysis shows that the proposed method can dramatically reduce calibration time compared to conventional autocalibrating parallel imaging methods, especially with large numbers of coil channels and synthesis patterns. Our invivo results on cardiac cine MRI also show the decoupling method preserves image quality of conventional kt GRAPPA and provides more accurate reconstruction and higher SNR than GRAPPA operator based methods due to its optimal combination of syntheses from different source data groups. The proposed data decoupling method is promising for online reconstruction for 3D cardiac cine MRI and other static 3D MRI applications.

Felix Breuer¹, Philipp Ehses^2,3, Nicole Seiberlich⁴, Martin Blaimer¹, Peter Jakob^1,5, and Mark Griswold^4
1Reserch Center Magnetic Resonance Bavaria, Würzburg, Germany, ²Max Planck Institute for Biological Cybernetics, Tübingen, Germany, ³Dept. for Neuroimaging, University Hospital Tübingen, Tübingen, Germany, Tübingen, Germany, ⁴Dept. of Radiology, Case Western Reserve University and University Hospitals of Cleveland, Cleveland, Ohio, United States, ⁵Experimental Physics 5, University of Würzburg, Würzburg, Germany

In this work an improved self-calibrating radial GRAPPA algorithm for real-time cardiac MRI is presented. It is shown that by subtracting the temporal average prior to the parallel MRI reconstruction significantly improved image quality can be achieved over pure radial GRAPPA. Real-time cardiac imaging data at a frame rate of 20fps in 2mm in-plane resolution are shown using only 16 radial projections per frame.

Mitchell A Cooper^1,2, Thanh D Nguyen², Pascal Spincemaille², Keigo Kawaji^1,2, Jonathan W Weinsaft³, Martin R Prince², 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

PROST (Parallel Reconstruction Observing Self consistency and Temporal smoothness) is a SPIRiT based paralell imaging technique implemented by extending the SPIRiT self-consistency kernel to the time domain. PROST is based on the assumption that each temporal phase has small changes with respect to its neighbors (temporal smoothness). PCA is incorporated into PROST to further express the information redundancy in the temporal dimension. PROST improves error performance compared to regular SPIRiT. When used for CINE imaging, PROST gives similar ejection fraction values compared to fully sampled data.

Ricardo Otazo¹, Cagdas Bilen², Yao Wang², Leon Axel¹, and Daniel K Sodickson^1
1Bernard and Irene Schwartz Center for Biomedical Imaging, NYU School of Medicine, New York, NY, United States, ²Department of Electrical Engineering, Polytechnic Institute of NYU, Brooklyn, NY, United States

Low-rank matrix completion is proposed as a new generalized approach to combining compressed sensing and parallel imaging by jointly exploiting implicit temporal and coil correlations without an explicit sparsifying transform or coil calibration procedure. A low-rank k-t matrix can be obtained by concatenating overlapping k-space blocks from consecutive time points and multiple coils to form the different columns. Reconstruction of k-t undersampled data is performed using an iterative singular value thresholding algorithm. We demonstrate the feasibility of reconstructing undersampled cardiac cine data.

Jun Liu¹, Jeremy Rapin¹, Ti-chiun Chang¹, Alban Lefebvre¹, Michael Zenge², Edgar Mueller², and Mariappan S. Nadar^1
1Siemens Corporate Research, Princeton, NJ, United States, ²Siemens AG, Healthcare Sector, Erlangen, Germany

High temporal resolution is often desired in Cardiac Magnetic Resonance Imaging (CMRI). Parallel imaging enables the reconstruction with a reduced the number of acquired frequencies, hence accelerating the acquisition, and spiral sampling scheme is widely used in CMRI. In this work, a new approach was proposed for dynamic cardiac MRI reconstruction by applying an L1 regularization based on the weighted 3D redundant Haar wavele for incorporating the smoothness in both spatial and temporal directions. Experiments conducted on a CMRI data with only 15 spokes for each temporal phase demonstrated the effectiveness of the proposed approach.

Irene Paola Ponce Garcia¹, Martin Blaimer², Felix Breuer², Peter M Jakob^1,2, and Peter Kellman^3
1Department of Experimental Physics 5, University of Würzburg, Würzburg, Bavaria, Germany, ²Research Center Magnetic Resonance Bavaria (MRB), Würzburg, Bavaria, Germany, ³Laboratory of Cardiac Energetics, National Institutes of Health, National Heart, Lung and Blood, Bethesda, Maryland, United States

In Dynamic Parallel Magnetic Resonance Imaging (pMRI) there are many reconstruction methods aiming to speed up acquisition and accurate reconstructions with high temporal and spatial resolution. It has been demonstrated that comparing to conventional k-t SENSE approach, Auto-Calibrating k-t SENSE saves acquisition time. Even though Auto-Calibrating k-t SENSE has been compared to conventional k-t SENSE, the comparison of the filter performing by both of them has not been jet evaluated. In this work, the performance in the temporal frequency of conventional and Auto-Calibrating k-t SENSE was evaluated using the 2-D Modulated Transfer Function (MTF).

Joshua D. Trzasko¹, Yunhong Shu², Armando Manduca¹, and Matt A Bernstein^2
1Physiology and Biomedical Engineering, Mayo Clinic, Rochester, MN, United States, ²Department of Radiology, Mayo Clinic, Rochester, MN, United States

Least-norm reconstruction of undersampled Cartesian MRI data, often referred to a “zero filling”, remains a popular strategy due to its simple and efficient implementation, and readily-characterized behavior. For non-Cartesian MRI, however, determination of an analogous reconstruction is computationally demanding and requires iterative methods that may be impractical for clinical use. In this work, we propose a novel and efficient numerical framework based on positive semi-definite constrained least-squares regression for generating accurate, non-iterative (i.e., direct) approximators of least-norm reconstructions of non-Cartesian MRI data.

Shuai Wang^1,2, Tian Liu³, Weiwei Chen⁴, Pascal Spincemaille², and Yi Wang^2,5
1University of Electronic Science and Technology of China, Cheng Du, Si Chuan, China, ²Radiology, Weill Cornell Medical College, New York, New York, United States, ³MedImageMetric LLC, New York, New York, United States, ⁴Radiology, Tongji Hospital, Tongji Medical College, Huazhong University of Science&Technology, Wuhan, Hubei, China, ⁵Biomedical Engineering, Cornell University, Ithaca, New York, United States

Several Quantitative Susceptibility Mapping (QSM) methods have been proposed to obtain tissue magnetic susceptibility in vivo. We compared various QSM methods in order to identify a robust QSM for clinical applications. We found that noise from the non-cone region in k-pace where the dipole inversion has been regarded as well behaved can contribute to substantial streaking artifacts in QSM. An accurate QSM needs address noise in the entire k-space, and not just the cone region. Evaluation on 21 consecutive patients indicated that the image space based morphology enabled dipole inversion is the most robust among three evaluated methods.

Yu Li¹, Feng Huang², Wei Lin², Randy Duensing², and Charles L. Dumoulin^1
1Imaging Research Center, Radiology Department, Cincinnati Children's Hospital Medical Center, Cincinnati, Ohio, United States, ²Invivo Diagnostic Imaging, Philips HealthCare, Gainesville, Florida, United States

The presented work aims to overcome the parallel imaging acceleration limit posed by spatial encoding capability of multi-channel coil sensitivity in clinical MRI. A previously reported framework of "correlation-based reconstruction" is used to convert high-speed imaging reconstruction to the estimation of correlation functions that may include multiple data correlation mechanisms underlying parallel acquisition. In the work presented here, we investigated whether coil sensitivity information and image content similarity can synergistically benefit correlation-based reconstruction for a static MRI scan.

Su-Chin Chiu¹, Hing-Chiu Chang², Tzu-Chao Chuang³, Fu-Nien Wang⁴, and Hsiao-Wen Chung^1
1Biomedical Electronics and Bioinformatics, National Taiwan University, Taipei, Taiwan, ²GE Healthcare, Taiwan, ³Electrical Engineering, National Sun Yat-Sen University, Taiwan, ⁴National Tsing Hua University, Taiwan

The FOV of Propeller MRI is approximately circular due to the rotating blade sampling pattern in the k-space, causing within-FOV aliasing artifact. In this article we proposed a self-training method similar to kt-Blast to restore the aliasing artifact to the conventional rectangular FOV. Phantom experiments were performed to verify the proposed method. Results show good agreement with conventional Cartesian sampling.

James H Holmes¹, Philip J Beatty^2,3, Howard A Rowley⁴, Zhiqiang Li⁵, and Jean H Brittain^1
1Global Applied Science Laboratory, GE Healthcare, Madison, WI, United States, ²Global Applied Science Laboratory, GE Healthcare, Thornhill, ON, Canada,³Physical Sciences, Sunnybrook Research Institute, Toronto, Ontario, Canada, ⁴Radiology, University of Wisconsin-Madison, Madison, WI, United States,⁵MR Engineering, GE Healthcare, Phoenix, AZ, United States

PROPELLER imaging with external calibration data driven parallel imaging acceleration has been shown to generate high T1 contrast while yielding sufficient blade width for robust motion correction. In this work, we present a crossed blade calibration method that enables greater imaging blade accelerations and decreased acquired echo trains for improved T1 contrast. Results from a PROPELLER acquisition with blades using echo trains of 7 and 3x effective acceleration for an effective blade width of 19 phase encodes are demonstrated to provide improved T1 contrast while maintaining wide blades for motion correction.

Naoharu Kobayashi¹, Djaudat Idiyatullin¹, Curtis A Corum¹, and Michael Garwood^1
1Center for Magnetic Resonance Research, University of Minnesota, Minneapolis, Minnesota, United States

Linogram sampling, introduced to radial MRI about 20 years ago, is a “semi-Cartesian” k-space sampling method, where the sampling pattern is a concentric square (2D) or cubic grid (3D). For linogram sampling, there is a corresponding reconstruction algorithm known as linogram reconstruction that does not need explicit interpolation and simplifies the Jacobian calculation, i.e. density correction. Therefore, linogram reconstruction has potential to improve resolution, minimize interpolation errors, and reduce computational time. In this study, we show point spread functions and reconstructed images from 3D linogram data that were calculated with the three reconstruction methods: gridding, backprojection and linogram reconstruction.

Sung-Hong Park¹, Jeffrey W. Barker^1,2, and Kyongtae Ty Bae^1
1Radiology, University of Pittsburgh, Pittsburgh, PA, United States, ²Bioengineering, University of Pittsburgh, Pittsburgh, PA, United States

Alternate Ascending/Descending Directional Navigation (ALADDIN) is a new imaging technique that provides interslice magnetization transfer (MT) asymmetry images with no separate preparation pulse. In this article, we investigated the ALADDIN MT asymmetry at various flip angles. The MT asymmetry signals peaked at around 60 degree. Centric phase encode order provided about twice higher MT asymmetry signals than linear phase encode order. Experimental results agreed with simulations results. The simulation results also showed that MT asymmetry signals decay with RF excitations faster at higher flip angle. The results of the study expand our knowledge on the interslice MT asymmetry signals.

Jochen Keupp¹, Osamu Togao², Jinyuan Zhou³, Yuriko Suzuki⁴, and Takashi Yoshiura^2
1Philips Research, Hamburg, Germany, ²Department of Clinical Radiology, Kyushu University, Fukuoka, Japan, ³Department of Radiology, Johns Hopkins University School of Medicine, Baltimore, MD, United States, ⁴Philips Electronics Japan, Tokyo, Japan

Amide proton transfer (APT) offers a novel route to sensitive MR molecular imaging of endogenous cytosolic proteins and is expected to play an important role in delineation/classification of active tumor tissue for therapy planning/monitoring. The length of RF saturation (T_sat) is a key for sensitivity, in particular at low protein concentrations. With novel parallel transmission based techniques, T_sat>1s is possible on clinical scanners. A thorough protocol optimization is needed for tumor contrast and contrast-to-noise ratio. In a phantom study and acquiring first exemplary clinical brain tumor data, the effect of T_sat on the APT contrast is studied in the range of T_sat=0.5s up to 4s.

Jyun-Ming Tsai¹, Teng-Yi Huang¹, Hsu-Hsia Peng², Yi-Chun Wu³, and Fu-Nien Wang^2
1Electrical Engineering, National Taiwan University of Science and Technology, Tapei, Taiwan, ²Biomedical Engineering and Environmental Sciences, National Tsing Hua University, Hsinchu, Taiwan, ³Molecular Imaging Center, Chang Gung Memorial Hospital, Taoyuan, Taiwan

This study presents an automatically real-time feedback optimization for frequency adjustment of magnetization transfer imaging. We proposed a general framework that can automatically optimize MT frequency based on the obtain image. The optimization can account for a whole field-of-view or a specific region-of-interest (ROI). We demonstrate this system by an experiment calibrating the central frequency of MT pulse for a ROI. The system was tested on a rabbit leg muscle. The frequency obtained by the real-time system was consistent with that obtained by a long sweep frequency scan. We therefore concludes the proposed system is a potential useful tool for MT-related studies, such as APT imaging.

Yee Kai Tee¹, Alexandr A Khrapichev², Manus J Donahue³, Nicola R Sibson², Stephen J Payne¹, and Michael A Chappell^1,4
1Department of Engineering Science, Institute of Biomedical Engineering, University of Oxford, Oxford, Oxfordshire, United Kingdom, ²Department of Oncology, CR-UK/MRC Gray Institute for Radiation Oncology & Biology, University of Oxford, Oxford, Oxfordshire, United Kingdom, ³School of Medicine, Vanderbilt University, Nashville, Tennessee, United States, ⁴Oxford Centre for Functional MRI of the Brain, University of Oxford, Oxford, Oxfordshire, United Kingdom

Discretizing the Gaussian pulses in pulsed CEST has been used recently as a way to calculate stimulated spectra. So far, different number of discrete intervals has been applied to perform the task. Since the number of intervals correlates directly to the computational time required, it is important to use the minimal discretization for efficient processing. In this study, simulations were used to calculate the minimal intervals required across a range of pulse parameters and the optimal values were then used to fit both in vitro and in vivo data. Excellent fits were found with minimum processing time.

Kenya Murase¹, and Shigeyoshi Saito^1
1Department of Medical Physics and Engineering, Osaka University Graduate School of Medicine, Suita, Osaka, Japan

We present a simple and fast method for solving the time-dependent Bloch equations in spin-locked chemical exchange saturation transfer (CESTrho) MRI using the 2-pool CEST model, and for calculating the longitudinal relaxation time in the rotating frame (T₁). When using the population-averaged longitudinal (R₁) and transverse relaxation rates (R₂), the T₁ values obtained by our method agreed with the approximate solutions given by Trott and Palmer, except for the on-resonance SL case. When the population-averaged R₁ and R₂ were not used, some differences were observed between them. Our method will be useful for better understanding and optimization of CESTrho MRI.

Monika Gloor¹, Klaus Scheffler^2,3, and Oliver Bieri^1
1University of Basel Hospital, Radiological Physics, Basel, Basel, Switzerland, ²MPI for Biological Cybernetics, MRC Department, Tübingen, Germany,³University of Tübingen, Neuroimaging and MR-Physics, Tübingen, Germany

An inversion recovery (IR) balanced steady-state free precession (bSSFP) sequence has been proposed for fast T1, T2, and spin density quantification. It has recently been shown that the presence of magnetization transfer (MT) effects on the transient phase of bSSFP experiments with short RF pulses leads to considerable deviations in calculated T1 and T2 values. In this work, an analytical expression of the two-pool IR bSSFP signal is presented taking MT effects into account. Numerical simulations of the Bloch equations are used to confirm the validity of the approximations made, and comparisons of the new equation to measurements are shown.

Zhongliang Zu^1,2, Vaibhav Janve^1,3, Junzhong Xu^1,2, Mark D Does^1,2, John Gore^1,2, and Daniel Gochberg^1,2
1Institute of Imaging Science, Vanderbilt University, Nashville, TN, United States, ²Department of Radiology, Vanderbilt University, Nashville, TN, United States, ³Department of Physics and Astronomy, Vanderbilt University, Nashville, TN, United States

APT imaging has shown promise in many clinical applications. However, the conventional CEST asymmetry analysis, by subtracting a label scan from a reference scan, is sensitive to confounding contributions from magnetic field inhomogeneities, inherently asymmetric macromolecular resonances, and lipid. In this study, we introduce a new CEST contrast that avoids these issues by creating label and reference scans based on varying the irradiation pulse turn angle (180 and 360 degree) instead of the frequency offset (3.5 and -3.5 ppm). Clear peak has been found at 3.5 ppm by using our method, but not by using the conventional asymmetry analysis.

Yee Kai Tee¹, Alexandr A Khrapichev², Nicola R Sibson², Stephen J Payne¹, and Michael A Chappell^1,3
1Department of Engineering Science, Institute of Biomedical Engineering, University of Oxford, Oxford, Oxfordshire, United Kingdom, ²Department of Oncology, CR-UK/MRC Gray Institute for Radiation Oncology & Biology, University of Oxford, Oxford, Oxfordshire, United Kingdom, ³Oxford Centre for Functional MRI of the Brain, University of Oxford, Oxford, Oxfordshire, United Kingdom

Pulsed CEST (PC) is the only feasible irradiation scheme for clinical application as a result of SAR and hardware limitations but there is no simple analytical solution available. Thus, it is typically treated as continuous CEST (CC) by finding its equivalent average field or power to exploit the closed-form analytical solution of CC. Although fitting the data by discretizing the short pulses used by PC is able to produce smaller fitted errors, there is no significant difference found for the important fitted parameters such as amide proton exchange rate (related to pH) when CC approximation, which is fast, is used.

Pouria Mossahebi¹, and Alexey A. Samsonov^2
1Biomedical Engineering, University of Wisconsin, Madison, WI, United States, ²Radiology, University of Wisconsin, Madison, WI, United States

We studied several ways to optimize efficiency of CRI at both 1.5 and 3T. We showed (for 3T case) that the CRI efficiency may be improved by including extra sample(s) from low Δ range (<1 kHz). Further, we demonstrated that SAR-efficient pulses may provide additional large efficiency gains in CRI. At the same time, high BW SAR-efficient pulses may lead to a significant error in qMT measures for the low offset frequency designs. One simple solution is to resort to longer SAR-efficient pulses.

Chao Xu^1,2, Christian Labadie^2,3, André Pampel², Jochen B. Fiebach¹, and Harald E. Möller^2
1Center for Stroke Research Berlin (CSB), Charite-Universitaetsmedizin Berlin, Berlin, Germany, ²Max Planck Institute for Human Cognitive and Brain Sciences, Leipzig, Germany, ³Laboratoire de Spectrométrie Ionique et Moléculaire, Université Claude-Bernard Lyon 1, Villeurbanne, France

Chemical exchange saturation transfer (CEST) permits the detection of dilute labile proteins to allow pH-weighted imaging. However, clinical applications of the technique are currently limited due to time and specific absorption rate (SAR) restrictions. We developed a three-dimensional (3D) pulse sequence with low SAR and embedded field map information. Potential application for detecting ischemic acidosis is demonstrated in a stroke patient imaged on the second day after symptom onset.

Pouria Mossahebi¹, and Alexey A Samsonov^2
1Biomedical Engineering, University of Wisconsin, Madison, WI, United States, ²Radiology, University of Wisconsin Madison, Madison, WI, United States

Accurate estimation of T1 values from VFA measurements requires consideration of magnetization transfer effects, which may introduce T1 bias roughly proportional to the macromolecular content. The fast T1 correction protocol developed in this work (VFA-MT) requires a single MT scan in addition to regular VFA measurements for accurate T1 mapping. The presented approach is promising for fast high-resolution whole brain T1 mapping within clinically acceptable scan times.

Zhongliang Zu^1,2, Ke Li^1,2, Richard Dortch^1,2, Seth Smith^1,2, Mark D Does^1,2, John Gore^1,2, and Daniel Gochberg^1,2
1Institute of Imaging Science, Vanderbilt University, Nashville, TN, United States, ²Department of Radiology, Vanderbilt University, Nashville, TN, United States

Conventional CEST sequences contain a long saturation block (several seconds of continuous wave or pulsed irradiation) before acquisition. To fill k-space, several saturation blocks followed by acquisitions may be required, especially in high-resolution imaging. The long scan time limits CEST imaging in clinical applications. Here we present a new CEST imaging method using interleaved balanced steady state free precession (SSFP) acquisitions (named SSFP-CEST) to shorten imaging time. Experiments on creatine/agar phantom show that the SSFP-CEST has similar contrast as that of pulsed-CEST, but the total acquisition time decreases by a factor of ~6.

Thomas Prasloski¹, Alexander Rauscher^2,3, Alex MacKay^1,2, Madeleine Hodgson¹, Irene Vavasour², Corree Laule⁴, and Burkhard Mädler^5
1Department of Physics and Astronomy, University of British Columbia, Vancouver, BC, Canada, ²Department of Radiology, University of British Columbia, ³UBC MRI Research Centre, University of British Columbia, ⁴Department of Pathology & Laboratory Medicine, University of British Columbia,⁵Department of Neurosurgery, University of Bonn

We present the application of an accelerated GRASE pulse sequence in the acquisition of data used to generate myelin water fraction maps. This sequence allows whole cerebrum coverage in scan times of less than 15 minutes. Further, the results of region of interest analysis indicate that myelin water fraction values correlate well with those of a previously validated sequence. Fast myelin water imaging could present new opportunities in both clinical and scientific research.

Ana-Maria Oros-Peusquens¹, Fabian Keil¹, Zaheer Abbas¹, Vincent Gras¹, Klaus H.M. Möllenhoff^1,2, and N. Jon Shah^1,3
1Institute of Neuroscience and Medicine - 4, Forschungszentrum Jülich, Jülich, Germany, ²Department of Cognitive Neuroscience, University Maastricht, Maastricht, Netherlands, ³JARA - Faculty of Medicine, RWTH Aachen University, Aachen, Germany

The method proposed here is based on a long-TR multiple-echo GRE measurement with 7:21 minutes acquisition time. It has the advantage that all B1 corrections, which are large, are multiplicative and can be approached heuristically, using a bias field correction. A simple exponential fit of the signal decay with echo time is the other postprocessing ingredient. Well-separated distributions are found for the WM and GM and agree very well with previous ones obtained at 1.5T. It is anticipated that this easy to use, high-quality water mapping method will open the road to routine, widespread quantitative MR imaging of water content.

Sean Deoni^1
1School of Engineering, Brown University, Providence, RI, United States

Rapid Myelin Water imaging can provide important information in de-myelinating disease (Multiple Sclerosis) as well as neuropsychological disorders that may be associated with “disconnectivity”. Although mcDESPOT is a rapid myelin water imaging, it suffers from an implicit assumption of two-component relaxation, which may not be true in voxels bordering CSF or edema. Here we introduce a third, non-exchanging component to address this deficiency. Results show improved myelin water estimation in areas bordering the ventricles and the periphery of the brain.

Klaus H. M. Möllenhoff^1,2, Fabian Keil¹, Zaheer Abbas¹, Vincent Gras¹, Miriam Rabea Kubach¹, Ana-Maria Oros-Peusquens¹, and N. Jon Shah^1,3
1Institute of Neuroscience and Medicine - 4, Forschungszentrum Jülich, Jülich, Germany, ²Department of Cognitive Neuroscience, University Maastricht, Maastricht, Netherlands, ³JARA - Faculty of Medicine, RWTH Aachen University, Aachen, Germany

Absolute water content in the human brain is highly regulated and shown to change in various diseases.Thus mapping of H₂O values in clinical acceptable measurement times is of great interest. Here we present a fast and accurate method for mapping H₂O values in vivo based on the 2D TAPIR approach.

Hye Young Heo^1,2, Brian J Dlouhy³, Daniel R Thedens¹, John Wemmie^3,4, and Vincent Magnotta^1,4
1Department of Radiology, University of Iowa, Iowa City, Iowa, United States, ²Department of Biomedical Engineering, University of Iowa, Iowa City, Iowa, United States, ³Department of Neurosurgery, University of Iowa, Iowa City, Iowa, United States, ⁴Department of Psychiatry, University of Iowa, Iowa City, Iowa, United States

The purpose of this study is to assess the ability of pH sensitive-T1ρ MRI to measure brain function. T1ρ contrast was compared with the BOLD contrast in the visual cortex during activation. In order to validate the specificity of T1ρ to pH changes, a sheep phantom study was performed, resulting in the double dissociation between T1ρ and T2* in acidified and oxygenated blood phantoms. In addition, we found that the release of the lactate by neural activity is consistent with the expected local acidosis.

Way Cherng Chen¹, Sean Foxley¹, and Karla Miller^1
1Centre for Functional MRI of the Brain (FMRIB), University of Oxford, Oxford, Oxfordshire, United Kingdom

Studies have shown that magnetic susceptibility introduces signal changes that are related to white matter tract structure. We previously presented a simple, small-scale 3D geometric model to estimate the microscopic field pattern that is driven by of various aspects of underlying tissue microstructure such as orientation to B0, myelin thickness and tissue iron concentration that can be used to predict susceptibility-based MR contrast. Here, a more robust and flexible 2D geometric model is introduced to allow more physiologically realistic representations of the actual WM microstructure.

Jason Su¹, Manojkumar Saranathan¹, and Brian Rutt^1
1Department of Radiology, Stanford University, Stanford, CA, United States

A view-sharing methodology in which higher order k-space samples are shared between data frames is applied along the flip angle dimension of a variable flip angle T1 mapping sequence to achieve an acceleration of 2.3x with nearly negligible loss in T1 estimation. The composite-derived maps had a percent error median of 0.104% and a 5-to-95 percentile range of -5.216% to 5.192%. Composite image quality and T1 map fidelity was dramatically improved by correcting mixed data with a scale factor based on values in the fully sampled center of k-space. The 5-to-95 percent error range shrank by a factor of 1.67.

Christopher Sica¹, and Christopher Collins^1
1Radiology, Pennsylvania State University, Hershey, PA, United States

A method for simultaneous B₁⁺ and T₁ mapping is presented. The proposed method acquires a series of spoiled 3D GRE scans with fixed TR and variable flip angle between scans. A spiral acquisition with off-resonance correction is utilized to speed acquisition. Non-linear fitting is applied to the signal expression M ₀ B₁^-(1-E₁)sin[ FA(λθ, ∆B0) ] / (1 – E₁cos[ FA(λθ, ∆B0) ]) to obtain B₁⁺ and T₁. This method is compared to an inversion recovery based method to measure T₁ in a set of doped water tubes, and agreement is on the order of several percent. In-vivo results from this technique are also presented.

Thanh D. Nguyen¹, Mitchell Cooper^1,2, Pascal Spincemaille¹, Ashish Raj¹, Yi Wang^1,2, and Susan A Gauthier^3
1Radiology, Weill Cornell Medical College, New York, NY, United States, ²Biomedical Engineering, Cornell University, ³Neurology and Neuroscience, Weill Cornell Medical College, New York, NY, United States

Conventional multi-component T2 relaxometry for myelin water fraction (MWF) mapping in the brain at 1.5T is time-consuming and has limited spatial coverage. The objective of this study is to develop a high resolution and high SNR 3D T2prep spiral protocol for whole brain myelin water mapping at 3T and to evaluate its performance in healthy volunteers by comparing with the conventional 1.5T acquisition. We found no significant changes in MWF between 1.5T and 3T. MWF maps obtained at 3T have lower coefficient of variance and better visualization of WM tracts.

Md. Nasir Uddin¹, R. Marc Lebel¹, and Alan H. Wilman^1
1Biomedical Engineering, University of Alberta, Edmonton, Alberta, Canada

Standard transverse relaxometry makes use of a multi-echo spin echo sequence, typically with a large number of echoes to enable accurate T2 fitting. Reducing the number of echoes would enable substantially reduced RF heating and increased slice coverage. However, in the presence of spatially varying RF fields, across or within a slice, standard exponential fitting methods are shown to perform poorly with limited echo trains. We demonstrate the use of a stimulated echo correction to enable reduction of echo trains to as few as four echoes while retaining accurate T2 fits in the human brain.

Ece Ercan¹, Peter Börnert², 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 Laboratories, Hamburg, Germany

We present a whole-brain tissue-based characterization of the ultrashort T2 components using 3D UTE MRI. Initial results show distinct distributions of ultrashort relaxation rates in gray and white matter. The estimated mean and distribution of the ultrashort T2 value for white matter is likely to contain contributions from cell membranes as well as from myelin, which is more ubiquitous in white matter than in gray matter. This implies that the actual ultrashort T2 value of myelin protons may be even shorter than the mean value found here and in previous works.

Jan Sedlacik¹, Kai Boelmans², Ulrike Löbel¹, Brigitte Holst¹, Susanne Siemonsen¹, and Jens Fiehler^1
1Neuroradiology, University Medical Center Hamburg-Eppendorf, Hamburg, Germany, ²Neurology, University Medical Center Hamburg-Eppendorf, Hamburg, Germany

MRI is able to quantitatively detect iron deposition by changes in transverse relaxation rates and signal phase. Recently, quantitative susceptibility mapping was used to further extract the underlying magnetic susceptibility from the MR signal phase information. Purpose of this study was to demonstrate that true quantitative iron content can be derived from the transverse relaxation rates by using the analytical model of the static spin dephasing regimen of spherical perturbations. The established control values of iron content in the normal aging brain will serve as control for future clinical studies of neurodegenerative diseases at our institution.

Thomas Neuberger^1,2, Tatjana Neuberger¹, Darin Clark³, Victoria Braithwaite^4,5, and Keith Cheng^3
1Huck Institutes of the Life Sciences, Pennsylvania State University, University Park, PA, United States, ²Department of Bioengineering, Pennsylvania State University, University Park, PA, United States, ³Department of Pathology, Penn State College of Medicine, Hershey, PA, United States, ⁴School of Forest Resources, Pennsylvania State University, University Park, PA, United States, ⁵Department of Biology, Pennsylvania State University, Univerity Park, PA, United States

The zebrafish is now widely recognized as a key model organism in biological research. Our study demonstrates that MRI can image the zebrafish in detail across its whole life span with an isotropic resolution of up to 8 microns. To perform these studies a standard bore 20 tesla system was utilized. Despite restricted space, we achieved a throughput of more than 60 specimens of early developmental stages. As the high resolution and contrast to noise ratio permits the identification of small details, we suggest that MRI could be a valuable tool to quantify the progression of disease in the zebrafish.

Ana-Maria Oros-Peusquens¹, Johannes Lindemeyer¹, Alard Roebroeck², Ralf A.W. Galuske³, Hansjuergen Bratzke⁴, and N. Jon Shah^1
1INM-4, Research Centre Juelich, Juelich, Germany, ²Dept. of Psychology, University of Maastricht, Netherlands, ³Dept. of Biology, TU Darmstadt, Germany, ⁴Dept. of Forensic Medicine, Faculty of Medicine, JWG-University, Frankfurt/M, Germany

We report on a multiparametric, high-resolution study of the motor cortex at 9.4T. Two intracortical bands were visualised, with contrast strikingly different from that of the surrounding cortex. We observe orientation dependence of the contrast in both magnitude and phase images as well as T2* and susceptibility maps. Interestingly, the dependence of the visualisation of the cortical bands on direction is complementary in magnitude and phase. This is an indication of microstructure dependence of T2* and phase contrast.

Ian J Rowland¹, McLean Gunderson², Seth Dailey², and Jose R Torrealba^3
1Medical Physics, University of Wisconsin-Madison, Madison, Wisconsin, United States, ²Department of Surgery, University of Wisconsin-Madison, United States, ³Department of Pathology and Laboratory Medicine, University of Wisconsin-Madison

Unlike optical microscopy, MR microscopy has few stains available to enhance contrast between tissue types. Whilst a significant number of MR microscopy studies use gadolinium based contrast agents to enhance available signal, enhanced tissue contrast is not guaranteed. In this study, we demonstrate that potassium dichromate may be utilized as a general MR histology stain to enhance contrast between a range of tissue types. The tissue specific generation of paramagnetic chromium species is responsible for the observed MR imaging contrast. Despite tissue oxidation, this study also demonstrates that standard histological procedures are unaffected following the chromium based MR staining.

Steven Harris¹, Liya Wang², Jing Huang², Lei Zhou¹, Hui Mao², and Xiaoping Hu^1
1Department of Biomedical Engineering, Georgia Institute of Technology / Emory University, Atlanta, GA, United States, ²Department of Radiology, Emory University, Atlanta, GA, United States

Adiabatic preparation pulses and the failure of the adiabatic condition for spins diffusing near the iron oxide nanoparticles can be used to generate contrast that is linearly increasing with particle concentration. In this work, in vitro work is extended to an in vivo model of nanoparticle accumulation in the mouse liver. The adiabatic full passage contrast, as well as a magnetization transfer compensated approach, is shown to be linearly correlated with organ iron concentration by biochemical measurement. The low nanoparticle concentrations used in this study suggest the approach may be suitable for in vivo quantitative molecular imaging studies.

Artem Khmelinskii¹, Esben Plenge², Peter Kok^3,4, Dirk Poot², Oleh Dzyubachyk³, Charl P. Botha⁴, Ernst Suidgeest⁵, Wiro Niessen^2,6, Louise van der Weerd^5,7, Erik Meijering², and Boudewijn P. F. Lelieveldt^3,4
1Division of Image Processing, Dept. of Radiology, LUMC, Leiden, The Netherlands, Leiden, Zuid-Holland, Netherlands, ²BIGR, Depts. of Radiology and Medical Informatics, ErasmusMC, Rotterdam, Netherlands, ³Division of Image Processing, Dept. of Radiology, LUMC, Leiden, Netherlands, ⁴Data Visualization Group, Dept. of Mediamatics, TU Delft, Netherlands, ⁵Dept. of Radiology, Molecular Imaging Laboratories Leiden, Section Nuclear Medicine, LUMC, Leiden, Netherlands, ⁶Quantitative Imaging Group, Dept. of Imaging Science & Technology, TU Delft, Netherlands, ⁷Depts. of Anatomy & Embryology, LUMC, Leiden, Netherlands

We present a new approach for producing highly resolved, localized isotropic volumes-of-interest in whole-body mouse MRI. This enables interactive HR visualization and exploration of anatomical structures in MRI. The idea is similar to that of well-known web-based geographical maps, where it is possible from a global overview image to zoom in on a detail of interest. Such functionality is relevant in a biomedical setting when working with high-resolution volumetric data. Using the method presented in this paper, from a global LR image the user can interactively zoom in on a sub-volume of interest.

Congbo Cai¹, Ying Chen¹, Jing Li¹, Lin Chen¹, Shuhui Cai¹, Jianhui Zhong², and Zhong Chen^1
1Department of Electronics Science, Xiamen University, Xiamen, Fujian, China, ²Departments of Radiology and Physics and Astronomy, University of Rochester, Rochester, New York, United States

A new reconstruction algorithm based on generalized Fourier transform is explored for sequences with linear frequency-swept excitation. A hybrid-encoding sequence is tested. This algorithm improves the spatial resolution of the images while maintaining the excellent field perturbation immunity of the sequence. In vivo experiments are performed on coronal and sagittal planes of rat. Compared to the traditional gradient echo sequence and spin-echo EPI sequence, the artifacts induced by motion effects, susceptibility variation and Nyquist ghost are greatly relieved in the images produced by the hybrid-encoding sequence.

Andrew Melbourne^1
1University College London, London, United Kingdom

This work is a review of techniques for retrospective automatic image registration techniques in dynamic contrast enhanced (DCE) MRI. Due to the length of time required to obtain a DCE MRI series image registration techniques are used to correct subject movement between imaging volumes prior to physiological model-fitting to extract features of the enhancement pattern. These techniques are often modified for the DCE setting so that they separate motion artefacts from contrast enhancement. Due to the proliferation of techniques a review is worthwhile so that knowledge of the technical details of image registration can be made more widely available.

Elizabeth M Tunnicliffe¹, Martin J Graves^2,3, and Matthew D Robson^4
1AVIC, Nuffield Department of Clinical Medicine, University of Oxford, Oxford, United Kingdom, ²Department of Medical Physics & Clinical Engineering, Cambridge University Hospitals NHS Foundation Trust, Cambridge, United Kingdom, ³Department of Radiology, University of Cambridge, Cambridge, United Kingdom, ⁴OCMR, Department of Cardiovascular Medicine, University of Oxford, Oxford, United Kingdom

There are a multitude of different SNR methods in MRI. In this educational poster we provide an overview of different SNR methods and their applicability, including guidance on when different approaches may be most useful.

Ives R Levesque¹, Nikola Stikov², and G Bruce Pike^2
1Electrical Engineering, Stanford University, Stanford, CA, United States, ²Montreal Neurological Institute, McGill University, Montreal, QC, Canada

This educational poster reviews the basic concepts of magnetization transfer and methods for quantitative MT imaging (QMTI). Various off- and on-resonance methods are discussed, along with their strengths and limitations for in vivo applications. General findings in neurological and muskuloskeletal imaging will also be discussed.