Mapping Fat, Susceptibility & Fields
Wednesday 22 April 2009
Room 313BC 10:30-12:30


Richard W. Bowtell and Nadim J. Shah

10:30 459. Young Investigator Award Finalist: Robust Water/Fat Separation in the Presence of Large Field Inhomogeneities Using a Graph Cut Algorithm
    Diego Hernando1, Peter Kellman2, Justin Haldar1, Zhi-Pei Liang1
Electrical and Computer Engineering, University of Illinois at Urbana-Champaign, Urbana, IL, USA; 2NHLBI, National Institutes of Health, Bethesda, MD, USA
    Water/fat separation is a classical problem for in vivo MRI. Although many methods have been proposed, robust water/fat separation is still challenging, especially in the presence of large field inhomogeneities. This work tackles the problem using a statistically-motivated formulation which jointly estimates the complete field map and water/fat images. This formulation results in a difficult (high-dimensional and non-convex) minimization problem, which is solved using a novel graph cut algorithm. The proposed method has good theoretical properties and an efficient implementation. It has proven effective for characterizing intramyocardial fat, producing robust water/fat separation in cases containing large field inhomogeneities due to susceptibility effects and magnet imperfections.
10:50 460. Fast Dynamic Fat-Water Separation Using Shorter Spatial-Spectral Excitation and Novel Temporal Acquisition
    Jing Yuan1, Tzu-Cheng Chao2, Riad S. Ababneh3, Yi Tang1, Lawrence P. Panych1, Bruno Madore1
Department of Radiology, Brigham and Women's Hospital, Harvard Medical School, Boston, MA, USA; 2National Taiwan University, Taipei; 3Department of Physics, Yarmouk University, Irbid, Jordan
    We present a novel fat-water separation strategy that combines a short spatial-spectral (SPSP) excitation with a TE-modulation acquisition scheme. The hybrid strategy effectively combines their strengths, and mitigates each other’s weaknesses. A SPSP pulse as short as 2ms suppresses most fat signals, and residual fat signals are identified by a varying TE acquisition in each time frame. Such modulations essentially label the phase of fat signals, by which fat is identified and removed through temporal processing in reconstruction. Higher temporal resolution could be achieved than with a normal SPSP excitation or a multi-echo Dixon’s separation.
11:02 461. Magnitude Fitting Following Phase Sensitive Water-Fat Separation to Remove Effects of Phase Errors
    Huanzhou Yu1, Ann Shimakawa1, Scott B. Reeder2, Charles A. McKenzie3, Jean H. Brittain4
Applied Science Laboratory, GE Healthcare, Menlo Park, CA, USA; 2Departments of Radiology, Medical Physics, Biomedical Engineering and Medicine, University of Wisconsin, Madison, WI, USA; 3Department of Medical Biophysics, University of Western Ontario, London, ON, Canada; 4Applied Science Laboratory, GE Healthcare, Madison, WI, USA
    Multi-point water-fat separation techniques rely on different water-fat phase shifts at multiple echo times to estimate the Bo fieldmap, water and fat. By utilizing the fieldmap smoothness, water-fat ambiguity can be resolved. However, these methods may be sensitive to eddy currents induced phase errors. In this work, the conventional phase-sensitive water-fat separation (first step) is followed by a fitting algorithm based on magnitude images (second step) for “fine-tuning”. The second step relies on the results from the first step for initial conditioning. The two-step approach is effective at removing phase errors for applications in both qualitative and quantitative water-fat separation.
11:14 462. A Weighted Gradient Regularization Solution to the Inverse Problem from Magnetic Field to Susceptibility Maps (Magnetic Source MRI): Validation and Application to Iron Quantification in the Human Brain
    Ludovic de Rochefort1,2, Tian Liu1, Bryan Kressler1, Jian Liu1, Pascal Spincemaille1, Jianlin Wu3, Yi Wang1
Radiology, Weill Medical College of Cornell University, New York, NY, USA; 2LMN, MIRCen, I2BM, DSV, CEA, Fontenay-aux-roses, France; 3Radiology, The 1st Hospital of Dalian Medical University, Dalian, Liaoning Province, China
    A reconstruction technique is presented to extract susceptibility maps from phase and magnitude MR data. The linear problem is solved using least-square regularization based on two terms to define boundary conditions and preserve edges. The technique is validated on phantom and applied in human to quantify iron in cerebral hemorrhage.
11:26 463. Susceptibility Mapping in the Human Brain at 3 and 7T
    Samuel James Wharton1, Andreas Schäfer2, Richard Bowtell1
Sir Peter Mansfield Magnetic Resonance Centre, School of Physics and Astronomy, University of Nottingham, Nottingham, UK; 2Department of Neurophysics, Max Planck Institute for Human Cognitive and Brain Sciences, Leipzig, Germany
    Phase images generated using gradient echo techniques at high field strengths show excellent contrast related to the different magnetic susceptibilities of various brain tissues. However, extraction of accurate anatomical information from these images is made difficult by the non-local relationship between the field perturbation and associated susceptibility distribution. Here, we validate a Fourier-based method for calculating 3D susceptibility maps from phase data using a specially-constructed agar phantom containing doped inclusions of known susceptibility. Results produced by applying the method to measuring the susceptibility of the substantia nigra and red nuclei from data acquired at 3 and 7T are also discussed.
11:38 464. Improvements in Quantitative Magnetic Susceptibility Mapping at Using Additional Low Resolution Scans
    Maarten J. Versluis1,2, Matthias J.P. van Osch1,2, Mark A. van Buchem1,2, Andrew G. Webb1,2
Radiology, Leiden University Medical Center, Leiden, Netherlands; 2C.J. Gorter Center for high field MRI, Leiden University Medical Center, Leiden, Netherlands
    Phase images measured with gradient echo sequences are sensitive to changes in the local magnetic field and can be used to identify regions with different magnetic susceptibilities. We have used the Fourier based relation between phase and susceptibility to calculate susceptibility maps. Simulations of phase maps showed improvements of susceptibility maps by combining high and low resolution data with different orientations to the magnetic field. Preliminary results of in-vivo data showed that artifacts in magnetic susceptibility maps can be reduced considerably using this approach.
11:50 465. Quantitative Susceptibility Mapping of Human Brain by Inverting Local Magnetic Fields Measured at Multiple Small Angles
    Tian Liu1,2, Pascal Spincemaille2, Ludovic de Rochefort2, Martin Prince2, Yi Wang1,2
Biomedical Engineering, Cornell University, Ithaca, NY, USA; 2Radiology, Weill Cornell Medical College, New York, NY, USA
    Quantitative susceptibility mapping of the brain would be valuable for assessing iron or calcium deposits associated with neurodegenerative and ischemic diseases, and for quantifying deoxygenated venous blood in fMRI. In this study, Calculation of Susceptibility through Multiple Orientation Sampling (COSMOS) was adapted to quantitatively map susceptibility of the cerebral venous blood. Veins were well-distinguished from the surrounding tissues and oxygen saturation level was assessed.
12:02 466. The Dependence of Tissue Phase Contrast on Orientation Can Be Overcome by Quantitative Susceptibility Mapping
    Karin Shmueli1, Peter van Gelderen1, Brian Yao1, Jacco A. de Zwart1, Masaki Fukunaga1, Jeff H. Duyn1
Advanced MRI Section, Laboratory of Functional and Molecular Imaging, National Institute of Neurological Disorders and Stroke, National Institutes of Health, Bethesda, MD, USA
    Interpretation of phase images is confounded by the dependence of the contrast on the tissue orientation relative to the main magnetic field (B0). To mitigate this, inverse Fourier methods have been proposed to reconstruct the underlying tissue magnetic susceptibility from the phase data. To evaluate this approach, we assessed the similarity of susceptibility maps calculated from phase images of human brain sections acquired at 0° and 90° relative to B0. The phase contrast was often reversed by rotation whereas the susceptibility maps were mostly unaffected. This demonstrates that the susceptibility calculation overcomes the strong orientation dependence of the phase contrast.
12:14 467. Removing Background Phase Variations in Susceptibility Weighted Imaging Using a Fast, Forward-Field Calculation
    Jaladhar Neelavalli1, Yu-Chung Norman Cheng2, Jing Jiang3, Ewart Mark Haacke2
Biomedical Engineering, Wayne State University, Detroit, MI, USA; 2Academic Radiology, Wayne State University, Detroit, MI, USA; 3The MRI Institute for Biomedical Engineering, Detroit, MI, USA
    We present here a novel method for removing background field effects from SWI phase images. The method involves predicting the air-tissue interface geometry induced field deviation and, from it the phase, and removing its contribution form the collected SWI phase data. The resultant images are referred to as Geometry Dependent Artifact Corrected phase images (GDAC phase) which lead to significant improvement in the processed susceptibility weighted magnitude images.