Lars Mueller¹, Andreas Wetscherek¹, Jutta Janke¹, Bram Stieltjes², and Frederik Bernd Laun^1
1Medical Physics in Radiology, German Cancer Research Center, Heidelberg, Germany, ²Quantitative Imaging-based Disease Characterization, German Cancer Research Center, Heidelberg, Germany

A double diffusion weighted sequence was implemented and tested on yeast suspension phantoms with different yeast concentrations. The sequence allowed us to determine the apparent exchange rate, which is a measure for water exchange rates across cell membranes, and the filter efficiency. Higher yeast concentrations result in lower exchange rates and higher filter efficiencies. Both parameters are dependent on the age of the yeast suspension and are shown to be good indicators of the vitality of the yeast suspension.

Ofer Pasternak¹, Markus Nilsson², Yoram Cohen³, Evren Ozarslan¹, Hans Knutsson⁴, and Carl-Fredrik Westin^1
1Harvard Medical School, Boston, MA, United States, ²Lund University, Lund, Sweden, ³Tel Aviv University, Tel Aviv, Israel, ⁴Linköping University, Linköping, Sweden

While most current double PFG methods concentrate on parameterization of new types of anisotropies in gray matter, we provide here a simple and intuitive filtered tensor based framework to explore the utility of double PFG for mapping white matter. The preliminary results shown here suggest that using filtered tensors has the potential to resolve complex fiber architecture in an intuitive, signal driven way, which is different than the typically complex model based HARDI methods.

Manisha Aggarwal¹ and Susumu Mori^1
1Department of Radiology, Johns Hopkins University School of Medicine, Baltimore, MD, United States

We investigate the effect of gradient frequency on diffusion MR contrasts derived using oscillating diffusion-sensitizing gradients in the fixed human cerebellum. Using diffusion MRI data acquired at 11.7 T with pulsed gradient and modified trapezoid-cosine oscillating gradient waveforms (at frequencies of 67, 100, 150, and 200 Hz), we report the generation of unique layer-specific contrasts in the human cerebellum. Quantitative maps derived from the rate of change in ADC measurements with gradient frequency revealed distinct anatomical contrasts highlighting specific cerebellar layers, which were compared with histologically-stained sections of the same tissue. The results demonstrate the potential of oscillating-gradient diffusion MRI acquisitions to generate anatomical contrasts that are sensitive to the microstructure of cellular layers in the human cerebellar gray matter.

Lebina Shrestha Kakkar¹, David Atkinson², Rachel W Chan², and Ivana Drobnjak^1
1Center for Medical Image Computing, University College London, London, United Kingdom, ²Center for Medical Imaging, University College London, United Kingdom

Estimation of brain micro-structure parameters can be used to improve understanding of neuro-degeneration due to ageing or dementia. Recent studies show these estimations are possible using pulsed gradient spin echo sequences but are more accurately estimated using trapezoidal oscillating gradient spin echo (tOGSE) sequences. Here, we implement tOGSEs on a 3T MRI scanner and test its feasibility by imaging: (1)microcapillaries (2)an asparagus stem and (3)the corpus callosum of two healthy volunteers. Their fiber diameters were estimated using only 5 measurements with the three-stage fitting model procedure. Our results suggest good agreement with the ground truth and histology for our phantoms.

Junzhong Xu¹, Hua Li¹, Kevin D Harkins¹, Jingping Xie¹, Mark D Does¹, and John C Gore^1
1Institute of Imaging Science, Vanderbilt University, Nashville, Tennessee, United States

Due to relatively long diffusion times used, small axons are usually overestimated using PGSE-based models. In the current study, the OGSE method was implemented in fixed spinal cord measurements with much shorter effective diffusion times to enhance detection sensitivity to small axons. The results show that the fitted mean axon diameters are in very good agreement with histology.

Hua Li¹, John C. Gore¹, and Junzhong Xu^1
1Institute of Imaging Science, Vanderbilt University, Nashville, TN, United States

Current PGSE-based approaches usually require long scanning times and high q-values to detect small axons, which are limited by the long diffusion times and the significantly reduced sensitivity to small spatial scales. OGSE has been shown to achieve much shorter diffusion times and hence may be able to detect smaller axons with high sensitivity. We measure the inner diameters of hollow microcapillaries using OGSE methods. Two ADC values are sufficient to extract the cylinder diameters accurately. OGSE is insensitive to the intrinsic diffusion coefficient in the relatively low frequency range. We propose OGSE as a fast and robust method for mapping axon sizes non-invasively.

Dan Benjamini^1,2, Michal E Komlosh^1,3, Uri Nevo², and Peter J Basser^1
1NIH, Bethesda, MD, United States, ²Tel Aviv University, Tel Aviv, Israel, ³USUHS, Bethesda, MD, United States

Noninvasive estimation of a non-parametric axon diameter distribution (ADD) provides sub-voxel microstructural information of white matter tissue. ADD infers on the nerve conductivity, and is influenced by several diseases, such as amyotrophic lateral sclerosis and multiple sclerosis. A recently suggested 2D double pulsed-field gradient MRI experiment is used on a porcine spinal cord. A voxel-by-voxel non-parametric ADD estimation provides a sub-voxel resolution map. Distinct domains where the ADDs are similar are then found. ADD estimation might prove to have clinical MRI applications owing to its critical functional role in the central and peripheral nervous systems, in normal and abnormal tissue.

Gonzalo Sanguinetti¹, Matt G Hall², Daniel C Alexander², and Rachid Deriche^1
1Athena Project-Team, INRIA, Sophia-Antipolis, France, ²Centre for Medical Image Computing, Department of Computer Science, University College London, London, United Kingdom

We are interested in retrieving information about the axon diameter distributions in white matter fiber bundles using NMR, which are commonly modelled as ensembles of parallel cylinders. We add regularization to the 1D-SHORE basis which results in more stable characterization of diameter distributions. To validate the method, we simulate NMR signals using the open source toolkit CAMINO. The results illustrate the enhanced estimation accuracy given by the regularization and provide an alternative validation of the SHORE based method.

Alessandro Proverbio¹, Bernard M. Siow^2,3, Eleftheria Panagiotaki³, Samuel Walker-Samuel², Mark F. Lythgoe², Adam P. Gibson¹, and Daniel C. Alexander^3
1Medical Physics and Bioengineering, University College London, London, United Kingdom, ²Centre for Advanced Biomedical Imaging, Division of Medicine, University College London, London, United Kingdom, ³Centre for Medical Image Computing and Department of Computer Science, University College London, London, United Kingdom

Investigation of tissue microstructure with non-invasive histology is a developing research area. Diffusion MRI (dMRI) can estimate features of microstructural components such as cell cytoarchitecture. Here develop signal models for Square Wave Oscillating Gradient Spin Echo (SWOGSE) dMRI. This study is performed on fixed subcutaneous xenograft tumour samples grown in nude mice: 3 LS174T, and 3 SW1222 cell lines. Oscillating Gradient Spin Echo probes shorter length scales than standard PGSE. The more tissue-like compartment model appears to represent the signal better, suggesting the feasibility of a clinical measurement of cytological properties of the tissue.

Novena A Rangwala¹, David B Hackney¹, and David C Alsop^1
1Department of Radiology, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, Massachusetts, United States

In this theoretical study we have investigated the effect of axonal dispersion in white matter on the accuracy of physiological parameters estimated with a diffusion model incorporating axonal dispersion into AxCaliber. Diffusion signal decay was estimated with a characteristic dispersion angle of 6°, at diffusion times 20–1000 ms and two axonal diameter distributions, and fit to AxCaliber equations without angular dispersion. Results show that although AxCaliber can fit the signal decay curves effectively, the fitted parameters do not accurately match the assumed physiological parameters. These results strongly suggest that models of diffusion signal decay include axonal dispersion, even in white matter tracts with well-aligned fibers.

Aritrick Chatterjee¹, Geoff Watson², Esther Myint³, and Roger Bourne^1
1Faculty of Health Sciences, University of Sydney, Sydney, NSW, Australia, ²Tissue Pathology and Diagnostic Oncology, Royal Prince Alfred Hospital, Sydney, NSW, Australia, ³Laverty Pathology, Sydney, NSW, Australia

Reduced water ADC in cancer tissue is commonly attributed to “increased cellularity”, however, this explanation includes many assumptions about the diffusion properties of tissue. We hypothesise that the biophysical basis of reduced ADC in prostate cancer may instead lie in an increased partial volume of low diffusivity epithelial cells and loss of higher diffusivity stroma and lumen space. Partial volumes of epithelium, stroma, and lumen space were measured in histology images. There was a significant increase in epithelium volume and decrease in stroma and lumen space with Gleason grade. Predicted ADC changes explained about half the variation seen in vivo.

Gregory Lemberskiy^1,2, Els Fieremans¹, Thorsten Feiweier³, Leon Axel¹, and Dmitry S Novikov^1
1New York University School of Medicine, Bernard and Irene Schwartz Center for Biomedical Imaging, Department of Radiology, New York, NY, United States, ²Sackler Institute of Graduate Biomedical Sciences, New York University School of Medicine, New York, NY, United States, ³Siemens AG, Healthcare Sector, Erlangen, Germany

The random permeable barrier model (RPBM) utilizes time-dependent diffusion to quantify cell size and membrane permeability. Human in-vivo RPBM validation was performed on individuals that experienced calf muscle atrophy while immobilized in a non-weight bearing cast. The RPBM was used to quantify the difference of myofiber size in immobilized and control calf muscles. Muscle groups of immobilized calf muscles of all volunteers were significantly smaller than those of the control leg. Additionally, the RPBM revealed that certain muscle groups atrophied less due to degrees of freedom offered by the cast.

Julien Valette^1
1Molecular Imaging Research Center (MIRCen), Commissariat à l'Energie Atomique, Fontenay-aux-Roses, France

An original diffusion modeling approach is proposed to capture some features of brain cells complexity and heterogeneity, while parameterizing the model with only a small set of parameters based on cell morphometric statistics as derived from microscopy. We exemplify how this approach allows evaluating the effect of long-range cell morphology (i.e. cellular processes branching and length) on the apparent diffusion coefficient (ADC) of intracellular molecules, as might be measured by diffusion-weighted MRI/spectroscopy at long diffusion times Td.

Milica Medved¹, Shiyang Wang¹, Steffen Sammet¹, Ambereen Yousuf¹, Gregory S Karczmar¹, and Aytekin Oto^1
1Radiology, University of Chicago, Chicago, Illinois, United States

The non-mono-exponential nature of the DWI signal decay in the prostate is well established. Current advanced modeling methods use a two component model, and the data is fit for the fast (corresponding to the coherent motion in blood vessels) and slow (describing the inherent diffusion properties of the glandular tissue) components. Using a new implementation of the inverse Laplace transform, we present evidence that multiple slow components can be present in the decaying DWI signal. We demonstrate evidence for single-component diffusion in the bladder, and for presence of up to three components in the peripheral zone of the prostate.

Ahmad Raza Khan^1,2, Lindsey A Leigland³, Steven G Kohama¹, Anda Cornea¹, Sune Nørhøj Jespersen⁴, and Christopher D Kroenke^1,2
1Division of Neuroscience, Oregon National Primate Research Center, Oregon Health & Science University, Beaverton, Oregon, United States, ²Advanced Imaging Research Center, Oregon Health & Science University, Portland, Oregon, United States, ³Advanced Imaging Research Center, Oregon Health & Science University, Beaverton, Oregon, United States, ⁴Center of Functionally Integrative Neuroscience, Aarhus University Hospital, Aarhus, Denmark

Recently, two-dimensional structure tensor (ST) analysis has been applied to light microscopy images for the purpose of validating diffusion anisotropy measurements. The ST analysis strategy enables the microscopy data to be analyzed in a high-throughput manner, and hence is suitable for group-comparison studies. However, 2D analyses are restricted to directions parallel to the imaging plane. Here we show it is possible to extend ST analyses to 3D using serial image "stacks" acquired with confocal microscopy of nonhuman primate brain tissue. It is expected that this generalization of the ST analysis will extend its range of applications in future studies.

Daniele Perrone¹, Jan Aelterman¹, Ben Jeurissen², Aleksandra Pizurica¹, Wilfried Philips¹, and Jan Sijbers^2
1IPI-TELIN-IMINDS, University of Gent, Gent, East Flanders, Belgium, ²Vision Lab, University of Antwerp, Antwerp, Antwerp, Belgium

Diffusion Weighted Imaging (DWI) was introduced to explore the human connectome in vivo; although many fiber tractography (FT) algorithms exist, proving the effectiveness of their estimates is challenging. We present a biologically and physically realistic software phantom, with brain-like fibres configuration and images, fully tunable in terms of ‘simulated acquisition’ parameters: a realistic bench test for quantitative analyses of every DWI-related algorithm.

Yurui Gao¹, Ann Choe², Iwona Stepniewskwa³, Xia Li², and Adam W Anderson^4
1VUIIS, Vanderbilt University, Nashville, Tennessee, United States, ²VUIIS, Vanderbilt University, TN, United States, ³Psychology, Vanderbilt University, TN, United States, ⁴BME, Vanderbilt University, TN, United States

In previous studies, we validated DTI-tractography-derived connectivity by comparing with the histological ground truth. Because DTI fiber orientation estimation is the basis of tractography, juxtaposing the DTI fiber orientation distribution (dFOD) and histological FOD (hFOD) is critical to validation studies. However, obtaining the 3D hFOD is challenging. This study introduces 1) two methods to extract the hFOD from z-stack micrographs and 2) the procedure to compare dFOD with hFOD functions.

Elan J. Grossman^1,2, Ivan I. Kirov¹, Oded Gonen¹, Dmitry S. Novikov¹, Robert I. Grossman¹, Matilde Inglese³, and Els Fieremans^1
1Center for Biomedical Imaging, Department of Radiology, NYU School of Medicine, New York, New York, United States, ²Department of Physiology and Neuroscience, NYU School of Medicine, New York, New York, United States, ³Department of Neurology, Radiology, and Neuroscience, Mount Sinai School of Medicine, New York, New York, United States

The relationship between compartment-specific white matter tract integrity (WMTI) parameters from diffusion MRI and concentrations of neurochemicals was investigated in vivo using 1H-MRS on a cohort of patients with mild traumatic brain injury (MTBI). Results demonstrated significant correlations associating n-acetylaspartate (NAA) with those WMTI parameters that are affected by intra-axonal diffusion and axonal density, suggesting the validity of a two-compartment non-exchange model of intra- and extra-axonal diffusion in a single WM fiber bundle. Our results also indicate that NAA may influence downstream processes related to axonal shrinkage, degeneration, and/or loss, as well as plasma osmolarity and myelin synthesis in MTBI.

Mohammed Salman Shazeeb^1,2, Sivakumar Kandasamy², Stuart Howes², and George Pins^2
1Radiology, University of Massachusetts Medical School, Worcester, MA, United States, ²Biomedical Engineering, Worcester Polytechnic Institute, Worcester, MA, United States

Invivo evaluation of biomaterial implant remodeling involves surgical removal of the implant for subsequent histological assessment. This approach is often destructive and imposes practical limitations on how effectively these materials can be evaluated. MRI has the potential to non-invasively monitor the remodeling of collagen scaffolds. This study investigated the development of a model system to quantify MRI parameters with angiogenesis and cellular infiltration of implanted collagen scaffolds using MRI and histological techniques. Correlations between MRI and histological parameters demonstrated that MRI is sensitive to specific remodeling parameters which can aid in the design of robust biomaterial scaffolds for tissue regeneration.

Naoki Ohno¹, Tosiaki Miyati¹, Tomohiro Chigusa¹, Hikaru Usui¹, Noam Alperin², Shinnosuke Hiratsuka¹, Akihiro Asano¹, Hirohito Kan³, Satoshi Kobayashi¹, Toshifumi Gabata¹, and Osamu Matsui^1
1Kanazawa University, Kanazawa, Japan, ²University of Miami, FL, United States, ³Nagoya City University Hospital, Nagoya, Japan

To clarify relations and mechanisms among blood-perfusion, water-diffusion, water-fluctuation, and biomechanics of the intracranial tissue, we developed an original cranial phantom for magnetic resonance imaging. The new cranial phantom consisted of a high-density polypropylene filter with intra- and extra-filter spaces, and a capacitor space, which were filled with water at 17 degrees centigrade. These correspond to a brain parenchyma, artery and vein, and cerebrospinal fluid space, respectively. Our original phantom makes it possible to clarify relations and mechanisms among blood-perfusion, water-diffusion, water-fluctuation, and biomechanics of intracranial tissue. Perfusion, diffusion, fluctuation, and biomechanics of the intracranial tissue interact in diverse ways.

Choong Heon Lee^1,2, Stephen J Blackband^1,2, and Fernandez-Funez Pedro^1,3
1Neuroscience, University of Florida, Gainesville, FL, United States, ²McKnight Brain Institute, Gainesville, FL, United States, ³Neurology, University of Florida, FL, United States

Understanding the complex brain architecture and functional connectivity using MRI has been a major interest in neuroscience. Based on the capability of MR microscopy to visualize the mammalian and human cells, the desire to map the entire human brain circuit at the resolution high enough to visualize cell-level structures is becoming a reality. Studies of a simpler brain, i.e. Drosophila, are significant due to its high genetic tractability together with high degree of conservation with humans at the genetic and cellular levels. To visualize the neuronal circuitry in the brain, we acquired MRM of it at 10 µm isotropic resolution.

Jeremy J Flint^1,2, Kannan M Menon^2,3, Brian Hansen⁴, and Stephen J Blackband^1,5
1Neuroscience, University of Florida, Gainesville, Florida, United States, ²McKnight Brain Institute, University of Florida, Gainesville, Florida, United States,³Biomedical Engineering, University of Florida, Gainesville, Florida, United States, ⁴Center for Functionally Integrative Neuroscience, Aarhus University, Aarhus, Denmark, ⁵National High Magnetic Field Laboratory, Florida State University, Tallahassee, Florida, United States

Recent developments allowing direct visualization of cellular structure in excised mammalian tissue using MRI techniques have created a need for micro-perfusion equipment capable of operating in conjunction with high-field spectrometer hardware. This study details the design and fabrication of an in-bore oxygenator capable of offering precise control of perfusate properties including pH and dissolved gas (O₂ & CO₂) content. In addition, a stability assessment is conducted which compares diffusion signal reproducibility over time under conditions of constant perfusion as compared to a static control group. Findings indicate statistical equivalence at 13 of 14 time-points sampled over a 21 hour period.

Lauren Burcaw¹, Jelle Veraart², Dmitry S Novikov¹, and Els Fieremans^1
1New York University School of Medicine, Bernard and Irene Schwartz Center for Biomedical Imaging, Department of Radiology, New York, NY, United States, ²Vision Lab, Department of Physics, University of Antwerp, Antwerp, Belgium

We report on time-dependent diffusion in white We report on time-dependent diffusion in white matter as measured on three normal human volunteers over diffusion times ranging from 26 to 400 ms. A decrease in axial diffusivity with time is found indicating the presence of restrictions to diffusion along the axonal direction. We also note a decrease in radial diffusion with respect to time, however this decrease is less pronounced. Our results imply that axial and possibly radial time-dependent diffusion may be observable using clinical systems.

Narina Norddin^1,2, Ned Charles¹, Nyoman Kurniawan³, Gary Cowin³, Laurence Gluch⁴, Carl Power⁵, Geoffrey Watson⁶, and Roger Bourne^1
1Faculty of Health Sciences, University of Sydney, Camperdown, NSW, Australia, ²Kulliyyah of Allied Health Sciences, International Islamic University Malaysia, Kuantan, Pahang, Malaysia, ³University of Queensland, Brisbane, Queensland, Australia, ⁴The Strathfield Breast Centre, Strathfield, NSW, Australia, ⁵University of New South Wales, Sydney, NSW, Australia, ⁶Royal Prince Alfred Hospital, Sydney, NSW, Australia

Conventional MRI has high sensitivity for breast cancer detection but poor specificity.Addition of DWI to a breast exam may increase specificity to around 90%.However,the biophysical basis of changes in diffusion weighted contrast in the breast and other non-neural tissue remains poorly understood.The study described here investigates the microscopic diffusion properties of formalin fixed breast tissue.Breast tissue glandular epithelium is similar to prostate tissue epithelium in having a low ADC relative to adjacent tissue.Low ADC may be a distinctive and diagnostically useful feature of glandular epithelia,not only in prostate and breast,particularly considering that 80-90% of all cancers are of epithelial origin.

Rainer Schneider^1,2, Jens Haueisen², and Josef Pfeuffer^1
1MR Application Development, Siemens Healthcare, Erlangen, Bavaria, Germany, ²Institute of Biomedical Engineering and Informatics, TU Ilmenau, Ilmenau, Thuringia, Germany

An asymmetric design is introduced to 2-D spatially selective radiofrequency (2DRF) pulses based on an echo-planar trajectory. The proposed design allows a significant reduction of pulse durations maintaining the excitation quality. This can be done independent of potential pulse accelerations with parallel transmission. Excitation quality with different asymmetry factors was analyzed in a phantom study. Asymmetric 2DRF pulses were evaluated in human diffusion-weighted experiments. Linked to the resulting TE savings, asymmetric 2DRF pulses offered SNR gains of up to 25% compared to the conventional 2DRF designs.

Suchandrima Banerjee¹, Emine Ulku Saritas^2,3, Rachel Connett¹, and Ajit Shankaranarayanan^1
1Global Applied Science Laboratory, GE Healthcare, Menlo Park, California, United States, ²Department of Electrical and Electronics Engineering, Bilkent University, Bilknet, Ankara, Turkey, ³National Magnetic Resonance Research Center, Bilknet University, Bilknet, Ankara, Turkey

Reduced field-of-view (rFOV) single-shot echo-planar imaging (ssEPI) with 2D Echo-planar (EP) RF excitation can improve diffusion image quality and enable high resolution, by reducing distortion. However slice coverage can be limited because of periodic repetitions of the excitation sidelobes appearing along the slice direction. We propose to address this constraint by using a multiband refocusing scheme that will simultaneously refocus multiple excitation lobes. Signal from the simultaneously refocused slice locations can be separated using parallel imaging method. We demonstrate that by using multiband refocusing with 2D EPRF in ssEPI, slice coverage can be increased while preserving rFOV and fat suppression.

Sean Foxley¹, Saad Jbabdi¹, Stuart Clare¹, and Karla Miller^1
1FMRIB Centre, University of Oxford, Oxford, OXON, United Kingdom

Due to the comparably sized RF wavelength at 7T, human brain imaging suffers from a central ‘bright spot’. Here, we investigated the implementation of a multiple flip angle approach to account for this B1 dependent artifact. Data were acquired at 7T of post-mortem human brain using a diffusion weighted steady-state free precession pulse sequence. Two flip angles were chosen using a described optimization method and DTI data were acquired using both. Results demonstrate that the primary diffusion direction was more accurately estimated in gray matter where signal loss due to B1 decreases have made this otherwise difficult to detect.

Yuji Iwadate¹, Mitsuharu Miyoshi¹, and Hiroyuki Kabasawa^1
1Global MR Applications and Workflow, GE Healthcare Japan, Hino, Tokyo, Japan

In diffusion preparation methods, the k-space signal modulation by T1 recovery results in image degradation if its effect is not negligible. We developed 3D diffusion-prepared balanced steady-state free precession pulse sequence with a variable flip angle technique to correct for the T1 recovery effects during data acquisition. In both simulation and phantom scan, variable flip angle reduced edge enhancement effect when we used centric view ordering. This technique can lead to homogeneous 3D diffusion weighted image acquisition of subjects with short T1 recovery times.

Martin Ott¹, David Porter², Felix Breuer¹, David Grodzki², Martin Blaimer¹, Björn Heismann², and Peter Jakob^1,3
1MRB Forschungszentrum für Magnet-Resonanz-Bayern e.V., Würzburg, Bavaria, Germany, ²Siemens AG, Erlangen, Bavaria, Germany, ³Department of Experimental Physics 5, University of Wuerzburg, Würzburg, Germany

Diffusion-Weighted Imaging (DWI) is one of the loudest and most challenging sequence in terms of acoustic-noise due to the requirement of strong and fast switching gradients. In this work, among other gradient optimizations, a rs-EPI readout is modified to consequently minimize slew rates and hence acoustic-noise. Methodical modifications are presented as well as corresponding acoustic-noise measurements. A significant reduction in acoustic-noise of over 15dB(A) could be achieved. In-vivo imaging results are compared to standard methods like single-shot EPI.

Jiazheng Wang¹ and Yongchuan Lai^2
1STO-MR, GE Healthcare, Beijing, Beijing, China, ²GE Healthcare, Beijing, China

EPI based diffusion imaging is widely used yet suffers from geometric distortions, which become more pronounced in high resolution imaging. This work presents a method to implement 2D excitation for high-resolution EP-DWI with reduced phase encoding FOV, and thus to reduce the geometric distortion. Comparing to prior-art technique, the 2D excitation in this work is equipped with a pair of walking-saturations, which can improve the excitation profile in the phase encoding FOV direction and get more homogeneous imaging results. The proposed method is also shown to increase the max number of slices per TR comparing to traditional 2D excitation.

A. Alhamud¹, Aaron T. Hess², Paul A. Taylor^1,3, Ernesta M. Meintjes¹, and André J.W. van der Kouwe^4
1Human Biology,MRC/UCT Medical Imaging Research Unit, University of Cape Town, Cape Town, Western Cape, South Africa, ²University of Oxford, Oxford, United Kingdom, ³African Institute for Mathematical Sciences, Western Cape, South Africa, ⁴Massachusetts General Hospital, Charlestown, Massachusetts, United States

Diffusion tensor imaging (DTI) is characterized by a long acquisition time in which B0 homogeneity may change from one TR to another, or from one DTI scan to another due to many factors especially if head position changes (e.g. acquiring DTI in two different phase encoding directions ‘AP and PA’ ). Several studies have focused on evaluating the changes in B0 for fMRI, but this issue has been largely neglected for DTI. In this work, we present the first study to measure, report and correct the inhomogeneity in the static field with simultaneous real time shim correction,TR-by-TR, in real time.

Tetsuo Ogino^1,2, Toshiaki Miyati³, Marc Van Cauteren⁴, Tomoya Nakamura⁵, Taro Takahara⁶, and Yutaka Imai^6
1Healthcare department, Philips Electronics Japan, LTD, Minato-ku, Tokyo, Japan, ²division of Health Sciences, Kanazawa University, Kanazawa, Ishikawa, Japan, ³Kanazawa University, Kanazawa, Ishikawa, Japan, ⁴Philips Healthcare LTD, Tokyo, Japan, ⁵Tokai University Hospital, Kanagawa, Japan, ⁶Tokai University, Kanagawa, Japan

A second moment nulling on diffusion gradients combined with ECG triggering for in-vivo myocardial diffusion weighted imaging was investigated its clinical feasibility. Uniform DWI signal was obtained with b-factor = 500 s/mm2 and clinically acceptable image quality with moderate signal non-uniformity was achieved up to b=800 s/mm2.

Xiaodong Ma¹, Feng Huang², Zhe Zhang¹, Bida Zhang², Sheng Fang³, and Hua Guo^1
1Center for Biomedical Imaging Research, Department of Biomedical Engineering, School of Medicine, Tsinghua University, Beijing, Beijing, China, ²Philips Research China, Beijing, China, ³Institute of nuclear and new energy technology, Tsinghua University, Beijing, China

POCS based Inherent Correction of motion-induced phase Errors (POCS-ICE) is proposed for the reconstruction in multi-shot spiral DWI, which can automatically correct the phase variation among different excitations without calculating them directly. Invivo experiments show that DW images from POCS-ICE present higher SNR and less aliasing artifacts than SENSE+CG when a large shot number is used. Therefore, it can contribute to shortening the spiral readout duration such that high resolution diffusion weighted images can be achieved.

Zhe Zhang¹, Feng Huang², Bida Zhang³, Sheng Fang⁴, and Hua Guo^1
1Center for Biomedical Imaging Research, Department of Biomedical Enginnering, School of Medicine, Tsinghua University, Beijing, China, ²Philips Healthcare, Gainesville, FL, United States, ³Philips Research China, Beijing, China, ⁴Institute of Nuclear and New Energy Technology, Tsinghua University, Beijing, China

Multi-shot strategies have been used for improving spatial resolution of DWI. MUSE is a multi-shot EPI based DWI method using SENSE for self-navigating of each shot. When many shots are used to decrease the EPI distortion and blurring, high SENSE reduction factor will cause inaccurate phase estimation and thus ghost artifacts. In this work, Self-feeding MUSE is proposed for robust phase estimation and improved MUSE reconstruction, even if the number of shots is high.

Merry Mani¹, Mathews Jacob², Vincent Magnotta², and Jianhui Zhong^1
1University of Rochester, Rochester, NY, United States, ²University of Iowa, Iowa, United States

Applications such as the study of human brain connectivity and pre-surgical planning can greatly benefit from high angular and spatial resolution diffusion data. Non-Cartesian multi-shot diffusion imaging schemes can offer high spatial resolution diffusion images. Coupled with high angular resolution schemes, it can enable the reconstruction of high fidelity fiber tracts. However, non-Cartesian multi-shot image reconstruction can be prohibitively time-consuming because of the need to account for motion-induced phase terms in the reconstruction to avoid motion artifacts. A fast reconstruction scheme is proposed using a PCA-based approximation, to accelerate the motion-compensated reconstruction of high angular and spatial resolution diffusion data.

Hing-Chiu Chang¹, Shayan Guhaniyogi¹, and Nan-Kuei Chen^1
1Brain Imaging and Analysis Center, Duke University Medical Center, Durham, North Carolina, United States

Multiplexed sensitivity encoding (MUSE) can effectively remove aliasing artifact due to shot-to-shot phase inconsistencies without relying on navigator echo. However, the conventional partial Fourier reconstruction procedure included in the original MUSE is highly susceptible to motion-induced k-space energy peak displacement, and the imaging throughput of multi-shot MUSE is significantly lower than single-shot DWI. We first develop a novel adaptive partial Fourier reconstruction procedure capable of producing high-quality multi-shot DWI, even in the presence of motion-induced k-data energy displacement. Next, we generalize the MUSE reconstruction to accommodate multi-band data, so that high-throughput, high-resolution and high-quality DWI can all be simultaneously achieved.

Qiuting Wen^1,2, Douglas A.C. Kelley³, Suchandrima Banerjee⁴, Janine M. Lupo², Duan Xu², Christopher P. Hess², and Sarah J. Nelson^1,2
1Graduate Program in Bioengineering, University of California, San Francisco & Berkeley, San Francisco, CA, United States, ²Radiology and Biomedical Imaging, University of California, San Francisco, San Francisco, CA, United States, ³Global Applied Science Laboratory, GE healthcare, San Francisco, CA, United States, ⁴Global Applied Science Laboratory, GE healthcare, Menlo Park, CA, United States

The goal of this study was to demonstrate the feasibility of acquiring 7T data suitable for NODDI in under 6 minutes with a tailored multiband technique for routine clinical application. In a brain tumor patient NODDI provided unique contrast within the T2 lesion. Although NODDI does not directly model tumor characteristics, it demonstrates great potential in revealing underlying tissue components that are complimentary to FA and ADC. With the multiband EPI technique, we reduced the acquisition time of this two-shell diffusion sequence to 5.5 minutes, making it clinically feasible. In addition to NODDI, both diffusion tensor and tractography data can also be generated from this protocol.

Angus Z. Lau^1,2, Elizabeth M. Tunnicliffe¹, Robert Frost³, Peter J. Koopmans³, Damian J. Tyler^1,2, and Matthew D. Robson^1
1Department of Cardiovascular Medicine, University of Oxford, Oxford, United Kingdom, ²Department of Physiology, Anatomy, and Genetics, University of Oxford, Oxford, United Kingdom, ³FMRIB Centre, University of Oxford, Oxford, United Kingdom

Remodeling of cardiac fibre structure contributes to impaired cardiac function, and is implicated in the development of cardiac arrhythmias and sudden cardiac death. Diffusion tensor imaging (DTI) allows non-invasive detection of microstructure in vivo, but low SNR leads to long scan times, limiting the patient population which can benefit from such exams. In this abstract, we investigate the feasibility of using simultaneous multi-slice (SMS) excitation to accelerate human DTI of the heart. We demonstrate three-fold scan time reduction of cardiac diffusion measurements, and we anticipate this development will make it practical to incorporate DTI within a comprehensive clinical exam.

Hans Knutsson¹ and Carl-Fredrik Westin^2,3
1Biomedical Engineering, CMIV, Linköping, Östergötland, Sweden, ²Harvard Medical School, MA, United States, ³Linköping University, Östergötland, Sweden

We present a novel approach to determine a local q-space metric that is optimal from an information theoretic perspective with respect to the expected signal statistics. The obtained metric will then serve as a guide for the generation of specific q-space sample distributions. It should be noted that the approach differs significantly from the classical estimation theory approach, e.g. one based on Cramer-Rao bounds. The basis for finding the q-space metric is to compute the information gain from measuring at a second q-space location given that we already have information from a first location.

Alexandra Tobisch^1,2, Gabriel Varela³, Rüdiger Stirnberg¹, Hans Knutsson⁴, Thomas Schultz^2,5, Pablo Irarrázaval^3,6, and Tony Stöcker^1
1German Center for Neurodegenerative Diseases (DZNE), Bonn, Germany, ²University of Bonn, Bonn, Germany, ³Biomedical Imaging Center, Pontificia Universidad Católica de Chile, Santiago, Metropolitan District, Chile, ⁴Linköping University, Linköping, Sweden, ⁵MPI for Intelligent Systems, Tübingen, Germany, ⁶Department of Electrical Engineering, Pontificia Universidad Católica de Chile, Santiago, Metropolitan District, Chile

The Compressed Sensing (CS) technique accelerates Diffusion Spectrum Imaging (DSI) through sub-Nyquist sampling in q-space and subsequent nonlinear reconstruction of the diffusion propagator. State-of-the-art DSI approaches that exploit CS apply Cartesian undersampling patterns. Recently, a method was proposed to generate 3D non-Cartesian sample distributions that aim for isotropic sampling of q-space. This work compares the new scheme to standard Cartesian undersampling patterns in sparse reconstruction of simulated diffusion signals. The diffusion propagator and the corresponding orientation distribution function of the reconstruction are found to deviate less from the ground truth when using an isotropic q-space sample distribution.

Qiyuan Tian¹, Ariel Rokem², Brian L. Edlow³, Rebecca D. Folkerth⁴, and Jennifer A. McNab^5
1Department of Electrical Engineering, Stanford University, Stanford, CA, United States, ²Department of Psychology, Stanford University, Stanford, CA, United States, ³Department of Neurology, Massachusetts General Hospital, Boston, MA, United States, ⁴Department of Pathology, Brigham and Women's Hospital, Boston, MA, United States, ⁵Department of Radiology, Stanford University, Stanford, CA, United States

Diffusion spectrum imaging is an approach to characterizing complex tissue microstructure. Stronger gradients enable expanded q-space coverage, which improves the spin-displacement resolution but also increases the q-space sampling density requirements. Here, we show three datasets acquired on a whole, fixed, human brain acquired with 300mT/m maximum gradients. These data are used to examine the effects of q-space sampling density on the fidelity of the voxel-wise orientation distribution functions (ODFs). Specifically, we show there is trade-off between ODF sharpness and aliasing artifacts when sampling density is insufficient to capture the spin-displacement pattern.

Alexandru V Avram¹, Joelle E Sarlls², Peter J Basser¹, and Evren Ozarslan^3
1Section on Tissue Biophysics and Biomimetics, NICHD, National Institutes of Health, Bethesda, MD, United States, ²NINDS, National Institutes of Health, Bethesda, MD, United States, ³Department of Radiology, Brigham and Women's Hospital, Harvard Medical School, Boston, MA, United States

Conventional methods for mapping 3D diffusion orientation distribution functions (dODFs) of neuronal microstructure require long acquisitions and are prone to quantitation errors due to subject motion. In this study we describe a method for directly measuring in vivo 3D dODFs with a novel technique called Rotating Field Gradient (RFG) diffusion MRI, which applies sinusoidal diffusion gradients in arbitrarily oriented planes to directly sample the dODFs along the axis of rotation. RFG dMRI represents a fast, direct and model-free method for mapping 3D ODFs that could represent a powerful alternative to conventional pulse-field gradient diffusion MRI techniques.

Jonas Cordes¹, Peter Neher¹, Hans-Peter Meinzer¹, Bram Stieltjes², and Klaus Maier-Hein^1,2
1Medical and Biological Informatics, German Cancer Research Center (DKFZ), Heidelberg, Baden-Württemberg, Germany, ²Quantitative Image-based Disease Characterization, German Cancer Research Center (DKFZ), Heidelberg, Baden-Württemberg, Germany

Here we present a set of metrics and experiments that evaluate the performance of several single- and multi-shell high angular resolution diffusion imaging (HARDI) acquisition schemes. In particular, we assess the angular resolution, the peak accuracy, the peak precision as well as the distribution function accuracy and distribution function precision at several different fiber configurations. We show that low b-value acquisitions in a multi-shell configuration can reach similar performances as high b-value acquisitions on a single shell.

Matteo Bastiani¹, Silvia De Santis^2,3, Derek Jones^2,3, Yaniv Assaf⁴, and Alard Roebroeck^1
1Department of Cognitive Neuroscience, Maastricht University, Maastricht, Limburg, Netherlands, ²CUBRIC, School of Psychology, Cardiff University, Cardiff, United Kingdom, ³Neuroscience and Mental Health Research Institute, Cardiff University, Cardiff, United Kingdom, ⁴Department of Neurobiology, The George S. Wise Faculty of Life Sciences, Tel Aviv University, Tel Aviv, Israel

The composite hindered and restricted model of diffusion (CHARMED) typically requires multiple acquisitions with high b-values and high angular resolution, both of which are difficult to obtain with clinical MRI systems. The purpose of this study is to investigate the feasibility of probing white matter microstructure at high spatial resolution in vivo at 3T, combining protocol optimization and a high amplitude, high slew-rate gradient set. Different levels of trade-off between resolution, SNR and diffusion contrast are compared and the results are further compared to a benchmark dataset obtained from the Human Connectome Project (HCP).

Vincent Gras¹, Ezequiel Farrher¹, Farida Grinberg¹, and N Jon Shah^1,2
1INM-4, Research Centre Jülich GmbH, Jülich, Germany, ²Department of Neurology, RWTH Aachen University, Aachen, Germany

We investigate the applicability of the dual echo steady state pulse sequence with diffusion-weighting field gradients (DW-DESS) to characterize the anisotropic diffusion of water molecules in a dedicated anisotropic diffusion fibre phantom. The conventional diffusion tensor imaging (DTI) model is used for the data analysis and a comparison with the gold standard pulse sequence for DTI, namely the DW spin-echo EPI sequence, is carried out. A good agreement between both pulse sequences is demonstrated. Finally the feasibility of applying DW-DESS for high resolution DTI in vivo is discussed.

Yeeliang Thian¹, Matthew R. Orton^2,3, David J. Collins^2,3, James A. d'Arcy^2,3, Martin O. Leach^2,3, Thorsten Feiweier⁴, and Dow-Mu Koh^1
1Department of Radiology, Royal Marsden Hospital, Sutton, Surrey, United Kingdom, ²CR-UK and EPSRC Cancer Imaging Centre,Institute of Cancer Research UK, Surrey, United Kingdom, ³Royal Marsden Hospital, Surrey, United Kingdom, ⁴Siemens AG, Healthcare Sector, Erlangen, Germany

The intra-voxel incoherent motion (IVIM) model relates signal decay on diffusion-weighted MRI (DWI) to tissue characteristics of perfusion and diffusivity. We explored the effects of altering the diffusion time of DWI sequences on IVIM perfusion (f,D*) and diffusion (D) related parameters in abdominal organs of healthy volunteers. At longer diffusion times, the calculated perfusion fraction (f) was significantly increased while the effect on the D and D* parameters was more variable, depending on the organ under study. The diffusion time has a significant impact on measured IVIM parameters in abdominal organs and should be reported when employing the IVIM model.

Liang Li¹, Liang Wang¹, Hui Liu², Ming Deng¹, Zhao-Yan Feng¹, and Ji-Hong Liu^3
1Department of Radiology, Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology, wuhan, hubei, China, ²Siemens Healthcare, Shanghai, China, ³Department of Urology, Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology, wuhan, hubei, China

In this study, we investigate the use of readout-segmented echo planar imaging(rs-EPI) for diffusion imaging of the prostate. This result show the rs-EPI sequence will be less affected by both chemical shift and magnetic susceptibility resulting in higher quality images.

Shohei Miyazaki¹, Masato Uchikoshi², Thorsten Feiweier³, Andre de Oliveira³, Souichirou Tateishi¹, Hitoshi Matsui¹, Takashi Horinouchi¹, and Katsuyuki Nakanishi^1
1Osaka Medical Center for Cancer and Cardiovascular Diseases, Osaka, Osaka, Japan, ²SIEMENS JAPAN K.K., Tokyo, Japan, ³SIEMENS AG, Erlangen, Germany

We evaluated the effect of duration of the diffusion-encoding gradient (ƒÂ) and/or diffusion time interval (ƒ¢) in DW-MRI. DW-MRI were performed with various ƒÂ or ƒ¢ in healthy volunteers and patients with metastatic brain tumor, and the ADC and kurtosis were estimated. In our results, the kurtosis was affected by ƒ¢ more significant than ƒÂ, whereas there was no significant change in ADC against to both of ƒÂ and ƒ¢. Our results suggested that it might be important to optimize the ƒ¢, leading usefulness in clinical diagnosis and assessment of the treatment effect of brain disease using kurtosis.

Sjoerd B. Vos¹, Chantal M.W. Tax¹, Fredy Visser², Peter R. Luijten², Alexander Leemans¹, and Martijn Froeling^2
1Image Sciences Institute, University Medical Center Utrecht, Utrecht, Utrecht, Netherlands, ²Department of Radiology, University Medical Center Utrecht, Utrecht, Utrecht, Netherlands

B0-drift-induced global signal decrease in DWI data causes a significant influence on the estimation of diffusion parameters. Quantitative diffusion metrics (FA and mean kurtosis) as well as directional information (first eigenvectors in DTI) are affected by this artifactual signal decrease during the scan session. By interspersing the non-diffusion-weighted images throughout the session, the signal decrease can be estimated and compensated before processing, minimizing the detrimental effects on all further analyses.

Jack Harmer¹, Rachel W Chan², Christian T Stoeck³, Constantin V Deuster³, David Atkinson², and Sebastian Kozerke^1,3
1Division of Imaging Sciences & Biomedical Engineering, King's College London, London, United Kingdom, ²Centre for Medical Imaging, University College London, London, Greater London, United Kingdom, ³Institute for Biomedical Engineering, University and ETH Zurich, Zurich, Switzerland

Advances in diffusion acquisition schemes employing stimulated echo modes now allow diffusion tensor imaging of the beating heart. However these pulse sequences almost invariably use single shot EPI readout modules that are extremely sensitive to magnetic field inhomogeneities. We demonstrate the feasibility of acquiring fully sampled single-shot in-vivo cardiac DTI data at 3T, in order to maximise SNR, by employing off-resonance correction as part of the reconstruction process. We demonstrate that such correction can be used to accurately reconstruct the correct geometry in areas of severe geometric distortion and signal pileup, whilst still allowing diffusion tensors to be reconstructed.

Oleg Posnansky¹, Myung-Ho In¹, and Oliver Speck^1
1Department of Biomedical Magnetic Resonance, Institute of Experimental Physics, Otto-von-Guericke University, Magdeburg, Germany

In this study, we compare fractional anisotropy (FA) maps and DT (diffusion tensor) fiber tracks built after application of extended PSF, and Reverse Gradient Polarity distortion correction methods. The results demonstrate that the extended PSF method is very efficient for the correction of distortions in both DW-EPI data with opposite PE polarity, and thus it unveils properly positioned and fully recovered fine anatomical structures in the brain FA maps and fiber tracks. Due to its high accuracy extended PSF provides overwhelming correspondence between anatomy and connectivity of the tissue.

Alan Seth Barnett^1,2, Elizabeth Hutchinson^1,2, M Okan Irfanoglu^1,2, and Carlo Pierpaoli^1,2
1NICHD, NIH, Bethesda, Maryland, United States, ²CNRM, USUHS, Bethesda, MD, United States

Higher order correction is sometimes required for adequate correction of eddy current induced distortion of diffusion weighted MR Images. The work consists of a description of the correction method and an example of a data set that requires cubic correction.

Sangwoo Lee¹, Mitsuhiro Bekku², Jeong Min Lee³, Jeong Hee Yoon³, Yuji Iwadate⁴, and Hiroyuki Kabasawa^4
1Global MR Applications and Workflow, GE Healthcare, Seoul, Korea, ²MR Engineering, GE Healthcare, Hino-shi, Tokyo, Japan, ³Seoul National University Hospital, Seoul, Korea, ⁴Global MR Applications and Workflow, Hino-shi, Tokyo, Japan

Tissue movements from respiration, cardiac, and peristaltic motion during diffusion gradient can cause signal loss, which often leads to local overestimation of ADC map. To reduce motion induced signal loss, maximum intensity projection, weighted averaging or weighted multiplication methods have been proposed. These post processing methods are simple and efficient; however the final images cannot achieve pixel intensities higher than acquired raw images. To overcome such limitation, we propose a new image combining method with homomorphic filtering (H-MSW) and show the feasibility of the proposed via simulation and in vivo experiment.

Jonathan I. Sperl¹, Tim Sprenger^1,2, Ek T. Tan³, Vladimir Golkov^1,4, Marion I. Menzel¹, Christopher J. Hardy³, and Luca Marinelli^3
1GE Global Research, Munich, BY, Germany, ²IMETUM, Technical University Munich, Munich, BY, Germany, ³GE Global Research, Niskayuna, NY, United States, ⁴Computer Vision Group, Technical University Munich, Munich, BY, Germany

In Diffusion Kurtosis Imaging (DKI) the data is sampled in a series of concentric shells in the diffusion encoding space (q-space). This work proposes to randomly undersample this multi-shell data in q-space (i.e. to acquire fewer data points) and to exploit the 1D Fourier relation between single rays in q-space and in the reciprocal propagator space in order to reconstruct the missing points based on the principles of compressed sensing using a non-cartesian total variation regularization. The benefits of this approach in terms of stability and accuracy of the kurtosis tensor estimation are shown for a volunteer diffusion MR data set using undersampling factors up to R=2.

Florian Knoll¹, Jos'e G Raya¹, Rafael O Halloran², Steven Beate¹, Eric Sigmund¹, Roland Bammer², Tobias Block¹, Ricardo Otazo¹, and Daniel K Sodickson^1
1Bernard & Irene Schwartz Center for Biomedical Imaging, Department of Radiology, NYU School of Medicine, New York, New York, United States,²Radiology, Stanford University, Stanford, California, United States

DTI allows to obtain quantitative measurements of tissue microstructure that no other technique can reveal. A simple and well defined signal model exists for DTI, which makes it an ideal candidate for model-based methods. The goal of this study is to introduce a new combination of compressed sensing and model based reconstruction where the sparsifying transform is evaluated directly in the domain of the diffusion tensor. Experimental results for truly accelerated in-vivo imaging are shown for both brain an MSK applications which demonstrate excellent performance of the model based approach.

Feng Huang¹, Zhe Zhang², Randy Duensing¹, Bida Zhang³, and Hua Guo^2
1Philips Healthcare, Gainesville, FL, United States, ²Center for Biomedical Imaging Research, Department of Biomedical Engineering, School of Medicine, Tsinghua University, Beijing, China, ³Philips Research China, Beijing, China

Multi-shot EPI has been proposed for high spatial resolution DWI. Due to inter-shot motion, there are phase inconsistency among shots. Unlike the existing reconstruction methods for ms-EPI, we treat this ms-EPI as accelerated dyanmic imaging. Instead of correcting the phase difference, we use the low-rank property of the data set and parallel imaging to reconstruct the images of each shot. Experiments demonstrate that the proposed method can robustly reconstruct high quality DWI with up to 8-shot data set acquired by an 8-channel head coil.

Rok Berlot^1,2, Claudia Metzler-Baddeley³, Derek K Jones³, and Michael J O'Sullivan^1
1Department of Clinical Neuroscience, Institute of Psychiatry, King's College London, London, United Kingdom, ²Department of Neurology, University Medical Centre Ljubljana, Ljubljana, Slovenia, ³Cardiff University Brain Research Imaging Centre (CUBRIC), School of Psychology, and the Neuroscience and Mental Health Research Institute, Cardiff University, Cardiff, United Kingdom

Partial volume effects due to cerebrospinal-fluid contamination (CSFC) lead to artefacts in diffusion MRI studies in populations prone to atrophy. We explored the influence of CSFC on apparent microstructural differences in amnestic mild cognitive impairment (MCI). A post-acquisition method of Free Water Elimination was used to correct for CSFC. Comparisons between controls and MCI were performed at several spatial levels in uncorrected and corrected data. CSFC shifted the pattern of involvement of individual temporal association tracts. Tract-based spatial statistics were robust to CSFC with notable localised exceptions. A proportion of group difference in mean histogram metrics was driven by CSFC.

Samer Sarfaraz Merchant¹, Arnold David Gomez¹, and Edward Hsu^2
1BioEngineering, University of Utah, Salt Lake City, Utah, United States, ²University of Utah, BioEngineering, Salt Lake City, Utah, United States

Myofiber atlases have been constructed from cardiac DTI datasets using voxel-based statistics, which is prone to noise and omits obvious functional and structural interconnectivities of the organ. In the current work, parametric modeling and principal component analysis were performed to characterize and investigate the variability of myofiber structure in a group of 6 mouse hearts. Results show that the hearts can be well represented by a 16-term parametric function in prolate spheroidal space, and majority of the heart-to-heart variability exists concentrically in the transmural dimension. These findings are valuable for design and interpretation of DTI cardiac structural measurement and modeling.

Ping-Hong Yeh¹, Namgyun Lee², John Morissette³, Arman A. Taheri³, Li-Wei Kuo⁴, Fang-Cheng Yeh⁵, Erick Jorge Canales- Rodríguez⁶, Wei Lui³, John Ollinger³, Terrence Oakes³, Mark L. Ettenhofer⁷, and Gerard Riedy^3
1Henry Jackson Foundation for the Advancement of Military Medicine, Bethesda, MD, United States, ²Korea Basic Science Institute, Korea, ³National Capital Neuroimaging Consortium, Bethesda, MD, United States, ⁴National Health Research Institutes, Taiwan, ⁵Carnegie Mellon University, PA, United States, ⁶FIDMAG Research Foundation, Germanes Hospitalaries and CIBERSAM, Barcelona, Spain, ⁷7Uniformed Services University of the Health Sciences, MD, United States

Recent work using Compressed Sensing (CS) reconstruction shows promising in greatly reducing diffusion spectrum imaging (DSI) scan time without jeopardizing critical image information. We evaluate the performance of CS reconstruction using dictionary-based training coupled with the Focal Underdetermined System Solver (FOCUSS) algorithm and L2 regularization on undersampled human brain DSI data acquired by a clinical 3T MR scanner within an acceptable time frame (< 20 minutes).

Vladimir Golkov^1,2, Marion I. Menzel¹, Tim Sprenger^1,3, Axel Haase³, Daniel Cremers², and Jonathan I. Sperl^1
1Diagnostics & Biomedical Technologies - Europe, GE Global Research, Garching n. Munich, Germany, ²Department of Computer Science, Technische Universität München, Garching n. Munich, Germany, ³Institute of Medical Engineering, Technische Universität München, Garching n. Munich, Germany

Recently, Joint Reconstruction has been proposed by Haldar et al. for SNR enhancement of diffusion-weighted images (DWIs), performing edge-preserving denoising by imposing identical edge constraints to all DWIs. In this work, we propose Semi-Joint Reconstruction to allow individual edges for each DWI in order to account for the fact that distinct DWIs can look quite dissimilarly, whereas similar DWIs contain common edge structures. Individual edge maps necessitate edge map denoising, for which we use DWI similarity weightings, truncated singular value decomposition of DWIs, and shearlet-based edge detectors. Results are comparable to Joint Reconstruction, but more stable to regularization parameter choice.

Vladimir Golkov^1,2, Marion I. Menzel¹, Tim Sprenger^1,3, Mohamed Souiai², Axel Haase³, Daniel Cremers², and Jonathan I. Sperl^1
1Diagnostics & Biomedical Technologies - Europe, GE Global Research, Garching n. Munich, Germany, ²Department of Computer Science, Technische Universität München, Garching n. Munich, Germany, ³Institute of Medical Engineering, Technische Universität München, Garching n. Munich, Germany

Reconstruction of diffusion-weighted images (DWIs) in diffusion MRI is usually done independently for each DWI, without exploiting structural correlations between the DWIs. In this work, we propose direct reconstruction of the average diffusion propagator (directly from k-space data), taking advantage of the DWIs being linked together via their Fourier relationship with the average propagator space, while regularization using five-dimensional total generalized variation (TGV) along both image space and diffusion space is applied. The results demonstrate the ability of the method to reconstruct q-space-undersampled data in a compressed sensing framework, simultaneously denoising the data.

Shiou-Ping Lee¹, Chung-Ming Chen², Li-Chun Hsieh¹, Wing-Keung Cheung¹, Yu-Chiang Chen¹, and Ming-Chung Chou^3
1Department of Medical imaging, Far Eastern Memorial hospital, Banqiao, New Taipei city, Taiwan, ²Insitute of Biomedical Engineering, National Taiwan University, Taipei, Taiwan, ³Department of Medical Imaging and Radiological Sciences, Kaohsiung Medical University, Kaohsiung, Taiwan

Recently, there were a variety of algorithms proposed to reconstruct orientation distribution function based on high angular resolution diffusion imaging data in order to resolve multiple fiber orientations. However, it was found that some fiber orientations were likely lost when fibers intersected at a small angle. A previous study performed the super resolved - constrained spherical deconvolution to resolve multiple fiber orientations crossing at a smaller angle, but the method was only suitable for diffusion-weighted datasets acquired in a spherical coordinate, such as q-ball imaging. Other HARDI datasets, such as diffusion spectrum imaging, acquired in a Cartestian coordinate whose fiber orientation distribution could not be obtained by using super-CSD. Hence, the purpose of this study is to propose a post-processing procedure which is suitable for sharpening ODFs of different HARDI datasets and resolving multiple fiber orientations by using super-CSD.

Gernot Reishofer¹, Kristian Bredies², Karl Koschutnig³, Margit Jehna⁴, Christian Langkammer⁵, David Porter⁶, and Hannes Deutschmann^4
1Radiology, Medical University of Graz, Graz, Styria, Austria, ²Institute for Mathematics and Scientific Computing, Universtiy of Graz, Austria,³Psychology, Universtiy of Graz, Austria, ⁴Neuroradiology, Medical University of Graz, Austria, ⁵Neurology, Medical University of Graz, Austria, ⁶Siemens AG, Healthcare Sector, MR R&D, Germany

It has been shown recently, that spatially dependent regularization of the diffusion tensor applied on readout-segmented echo planar imaging (rs-EPI) with 2D navigator-based reacquisition significantly improves fractional anisotropy (FA) maps and tractography. In this work we propose a novel approach for automatic regularizing the entire diffusion tensor utilizing a three dimensional implementation of total generalized variation (TGV). The evaluation of the noise distribution of the diffusion tensor by means of ICA allows for an automatic update of the regularization parameter making the proposed algorithm user-independent. Furthermore the incorporation of the locally varying noise distribution allows for a spatially dependent regularization.

Alberto Ciarmatori^1,2, Elisa Cenacchi¹, Tiziana Costi¹, and Luca Nocetti^1
1Medical Physics Department, University Hospital "Policlinico", Modena, Italy, ²Medical Physics School, University "Alma Mater Studiorum", Bologna, Italy

This study focused on separation of vascular and non-vascular component of DWI signal decay. Brain scans were performed using a 3T whole body MR on six healthy volunteers. DWIs with multiple b-values (0-5-10-20-30-40-50-75-100-150-200-300-500-1000 s/mm2) were obtained. Datas were fitted with a biexponential function with Diffusion Coefficient and Perfusion Fraction obtained from high-b (>250 s/mm2) monoexponential fit. A temporal optimized (6 min) sequence was found . Significative differences were observed D and standard ADC as well as between gray matter and white matter diffusion parameters. In particular D* of gray matter resulted significantly greater than in white matter.

S. Tantisatirapong¹, N. P. Davies^1,2, D. Rodriguez³, L. Abernethy⁴, D. P. Auer^3,5, C. A. Clark^6,7, R. Grundy^3,5, T. Jaspan⁵, D. Hargrave⁷, L. MacPherson², M. O. Leach⁸, G. S. Payne⁸, B. L. Pizer⁴, S. Bailey⁹, A. C. Peet^1,10, and T. N. Arvanitis^10,11
1University of Birmingham, Edgbaston, Birmingham, United Kingdom, ²University Hospitals Birmingham NHS Foundation Trust, Birmingham, United Kingdom, ³University of Nottingham, Nottingham, United Kingdom, ⁴Alder Hey Children’s NHS Foundation Trust, Liverpool, United Kingdom, ⁵University Hospital Nottingham, Nottingham, United Kingdom, ⁶University College London, London, United Kingdom, ⁷Great Ormond Street Hospital, London, United Kingdom, ⁸The Institute of Cancer Research and Royal Marsden Hospital, Sutton, United Kingdom, ⁹Northern Institute for Cancer Research, Newcastle University, Newcastle upon Tyne, United Kingdom, ¹⁰Birmingham Children’s Hospital NHS Foundation Trust, Birmingham, United Kingdom, ¹¹Institute of Digital Healthcare, WMG, University of Warwick, Coventry, United Kingdom

This paper presents integration of multimodal MR image based texture analysis which take advantage of complementary information derived from structural and diffusion MR images. A supervised machine learning approach is used to achieve image analysis based on textural features from individual image types of T2, T1-post contrast and ADC, as well their combination, in order to characterize the multicenter dataset of the most common pediatric brain tumors; medulloblastomas (MB), pilocytic astrocytomas (PA), and ependymomas (EP).

Harini Veeraraghavan¹, Joseph O. Deasy¹, and Richard K.G. Do^1
1Memorial Sloan Kettering Cancer Center, New York, New York, United States

Bulk motion artifacts adversely impact the reproducible measurement of ADCs in the liver. We present a combined segmentation and sequential registration to align multi b-valued DWI images that removes artifacts and computes repeatable ADC mappings. Our approach uses the automatic volumetric segmentation of structures of interest (tumor) with a sequential registration such that the artifacts are most removed around those structures. Evaluation over multiple trials with introduced motion on 14 patients show that our approach results in least artifacts and lowest standard deviation of mean ADC in the tumors compared to either no registration or simultaneous affine registration.

Susanne Will¹, Petros Martirosian¹, Christina Schraml^1,2, Bernd Kardatzki³, Michael Erb³, Günter Steidle¹, and Fritz Schick^1
1Department of Diagnostic and Interventional Radiology, University of Tuebingen, Tuebingen, Germany, ²Section on Experimental Radiology, University of Tuebingen, Germany, ³Department of Biomedical Magnetic Resonance, University of Tuebingen, Germany

Classification of anatomical renal structures in diffusion-weighted echo-planar images may be difficult to perform. For this reason a simple transfer of segmented areas from high-resolution FSE/GRE images with suitable visualization of the anatomy to DTI images would be helpful but is hampered by geometric distortions. The purpose of this study was to assess the feasibility of renal segmentation and their internal structures in distortion corrected DTI images. Six volunteers were examined at 3T. It was found that DTI values can be accurately determined in specific areas of distortion corrected images by manual renal segmentation in corresponding high-resolution anatomical images.

Hedok Lee^1,2, Joe Cavallo¹, Vivek Goyal¹, Sagar Dhandia¹, Michael Cutrone¹, Elyssa Chen¹, Tian Fang¹, Helene Benveniste^1,2, and Thomas Floyd^1,2
1Anesthesiology, State University of New York at Stony Brook, Stony Brook, New York, United States, ²Radiology, State University of New York at Stony Brook, Stony Brook, New York, United States

In this study we compared manual and semi-automated computerized anatomical delineation techniques in ex-vivo diffusion tensor imaging of the mouse brain. The accuracy of both FAs and anatomical volumes in the computerized technique was comparable or just below the discrepancies caused by subjective bias between the two raters. We conclude that the computerized method is effective and time efficient method in analyzing FA and anatomical volumes.

Tsang-Chu Yu¹, Yi-Ping Chao¹, Li-Wei Kuo², Chung-Chih Lin¹, Shih-Yen Lin^2,3, Hengtai Jan², Claudia Metzler-Baddeley⁴, and Derek Jones^4
1Department of Computer Science and Information Engineering, Chang Gung University, Taoyuan, Taiwan, ²Institute of Biomedical Engineering and Nanomedicine, National Health Research Institutes, Miaoli, Taiwan, ³Department of Computer Science, National Chiao Tung University, Hsinchu, Taiwan,⁴School of Psychology, Cardiff University, Cardiff, United Kingdom

The main purpose of this study is to develop a graphics processing unit based framework for brain network analysis that permit networks comprising much larger numbers of nodes and provide the acceleration for processing. From the results, our implementation for the calculation of all pair shortest paths could reduce half of time with brain connectivity toolbox (BCT) and 638x speedup with Gretna in simulation random network with larger number of nodes (>8k). Moreover, our algorithm also shows better performance in human brain data with 1.37x and 21x speedup in comparison with BCT and Gretna respectively.

Hsu Yung-Chin¹ and Tseng Isaac Wen-Yih^1,2
1Center for Optoelectronic Medicine, College of Medicine, National Taiwan University, Taipei, Taipei, Taiwan, ²Graduate Institute of Brain and Mind Sciences, College of Medicine, National Taiwan University, Taipei, Taiwan

In the study, we propose a new algorithm to iteratively estimate the mean shape of a group of tract bundles. The basic idea of our algorithm is: performing registration on tract bundles and taking the average on orientation maps. The conversion from orientation maps to tract bundles is achieved via fiber-tracking algorithms, and the reverse way is via modeling the tract bundles as currents. The algorithm is demonstrated on ten right cingulum bundles, and the results show that this method could properly estimate the mean cingulum bundle. This method has the potential to facilitate computational anatomy of the brain.

Koji Sakai¹, Ryusuke Nakai¹, Kei Yamada², Jun Tazoe², Kentaro Aakazawa², and Naozo Sugimoto^1
1Kyoto University, Kyoto, Kyoto, Japan, ²Kyoto Prefectural University of Medicine, Kyoto, Japan

Among MR thermometry, the most clinically applicable may be the post processing of diffusion-weighted images (DWI). However, this DWI-based method might be influenced by the composition of CSF, which can strongly affect its viscosity and diffusivity. The purpose of this study was to investigate the influence of protein content on DWI thermometry by using temperature-controlled artificial CSF (ACSF). The results of DWI thermometry were not influenced by the protein content in the range of 0.0 to 8.01 mg/ml (approximately 40 times protein content of normal adult) at near body temperature.

Ryan Hutten¹, Nisa Desai¹, Demetrius Maraganore^2,3, Robert R. Edelman^1,4, and Ying Wu^1,5
1Radiology, Northshore University Health System, Evanston, IL, United States, ²Neurology, Northshore University Health System, IL, United States,³Neurology, University of Chicago, IL, United States, ⁴Radiology, Northwestern University Feinberg School of Medicine, Chicago, IL, United States,⁵Radiology, University of Chicago, Chicago, IL, United States

Automated tract based diffusion tensor imaging (DTI) and magnetization transfer ratio (MTR) of the nigro-striatal tract (NST) represents a high innovative neuroimaging approach in Parkinson’s disease. The NST is a major dopamine pathway that is critically involved in Parkinson’s Disease (PD) diagnosis and treatment modulation. The development of the automated NST overcomes the challenges associated with segmenting SN, and the necessity of manual seed regions for tractography. We demonstrate considerably improved reliability that eliminates operator-induced error that will benefit future longitudinal studies of Parkinson's disease.

Fang-Yu Nien¹, Seong Yong Lim¹, Yu-Han Hong¹, Ya-Han Chang¹, and Jun-Cheng Weng^1,2
1School of Medical Imaging and Radiological Sciences, Chung Shan Medical University, Taichung, Taiwan, ²Department of Medical Imaging, Chung Shan Medical University Hospital, Taichung, Taiwan

Characterizing complex anatomy at different stages of brain development not only aids in understanding this highly ordered process but also provides clues to detect abnormalities caused by genetic or environmental factors. Diffusion MRI allows for the in vivo delineation of white matter tracts in the brain in a manner that is individualized to the particular neuroanatomy of each subject. Diffusion anisotropy and diffusivity change in some brain regions with demyelinating disease and also with neural development. Previously we used diffusion tensor imaging (DTI) to detect the changes of white matter tracts of rabbit during brain development. However, DTI suffers from resolving the complicated neural structure, i.e. fiber crossing, which is observed frequently during mature period. Therefore the goal of this study was to characterize the changes of quantitative diffusion indices in the developing rabbit brains using generalized q-sampling imaging (GQI).

Chantal M.W. Tax¹, Sjoerd B. Vos¹, Max A. Viergever¹, Martijn Froeling², and Alexander Leemans^1
1Image Sciences Institute, University Medical Center Utrecht, Utrecht, Utrecht, Netherlands, ²Department of Radiology, University Medical Center Utrecht, Utrecht, Utrecht, Netherlands

Interpolation in q- or b-space can be used to convert between different sampling patterns (e.g. from multi shell to grid as in hybrid diffusion imaging) to be able to compare between diffusion MRI reconstruction methods, or to detect and correct for artifacts, amongst others. In this work we perform a comparison between different interpolation methods (i.e. nearest neighbor, linear, cubic and model based) in q- and b-space, on simulated data.

Masaaki Hori¹, Koji Kamagata¹, Mariko Yoshida¹, Nozomi Hamasaki², Shuji Sato², Kouhei Kamiya^1,3, Yuriko Suzuki⁴, Michimasa Suzuki¹, and Shigeki Aoki^1
1Radiology, Juntendo University School of Medicine, Tokyo, Japan, ²Juntendo University Hospital, Tokyo, Japan, ³Radiology, The University of Tokyo, Tokyo, Japan, ⁴Philips Electronics Japan, Ltd, Tokyo, Japan

We investigate two-component low q-space diffusion MRI in evaluation of microstructural changes in the spinal cord in patients with cervical spondylosis in vivo, as a feasibility study. Five patients were enrolled in this study. There were significant differences in root mean square displacement of water molecules in extra- and intra- cellular spaces between compressed and uncompressed area in the spinal cords. More studies of the imaging pathological correlation and improvement of image quality are needed; this technique has the potential to provide new information in patients with spinal cord in vivo.

Emmanuel Caruyer¹ and Ragini Verma^1
1Section of Biomedical Image Analysis, Department of Radiology, University of Pennsylvania, Philadelphia, PA, United States

We present a novel family of rotational-invariant measures for High Angular Resolution Diffusion Imaging1 (HARDI) data. This research is motivated by need for the developing new biomarkers and thereby facilitating population-based studies in HARDI. From a spherical harmonics (SH) representation of the angular information in diffusion MRI, we derive 12 rotational-invariant measures computed as homogeneous polynomials of the SH coefficients. These measures are expected to be sensitive to subtle changes in white matter structure, not captured by FA or GFA. The large number of measures will be used to create population classifiers, paving the way for pathology-specific biomarkers.

Andrew R. Hoy^1,2, Chen Guan Koay¹, Steven R. Kecskemeti^2,3, and Andrew L. Alexander^1,2
1Medical Physics, University of Wisconsin, Madison, Wisconsin, United States, ²Waisman Laboratory for Brain Imaging and Behavior, Madison, Wisconsin, United States, ³Radiology, University of Wisconsin, Madison, Wisconsin, United States

Diffusion tensor imaging yields information about tissue microstructure. However, when a single voxel contains tissue and free water, DTI is not appropriate. A two-tensor fast diffusion estimation model has been proposed to correct this shortcoming. This model was implemented in a novel manner, and the acquisition parameters optimized through Monte Carlo simulations. The optimal acquisition with 68 diffusion-weighted encoded images had three diffusion-weighted shells (b-value in s/mm2 x number of directions) of 200x12, 650x40, 1500x12. This was confirmed in vivo. The model is useful for tissues adjacent to CSF and removing artifacts from CSF blurring and ghosting.

Yi Sui^1,2, Frederick C. Damen^1,3, and Xiaohong Joe Zhou^1,4
1Center for MR Research, University of Illinois at Chicago, Chicago, Illinois, United States, ²Department of Bioengineering, University of Illinois at Chicago, Chicago, Illinois, United States, ³Department of Radiology, University of Illinois Hospital & Health Sciences System, Chicago, Illinois, United States,⁴Departments of Radiology, Neurosurgery and Bioengineering, University of Illinois Hospital & Health Sciences System, Chicago, Illinois, United States

An integrated high-resolution, multi-b-value diffusion acquisition approach has been developed with reduced FOV and phase-sensitive signal averaging. Using this approach, distortion-free brain stem diffusion images with an in-plane resolution of ~0.6×0.6 mm² have been obtained at b-values up to 4000 s/mm² and successfully fitted to a non-Gaussian diffusion model (Fractional Order Calculus model). This study demonstrates the potential for extending the applications of high b-value non-Gaussian diffusion models to fine brain structures.

Louise Munk Rydtoft¹, Brian Hansen¹, Mikkel Bo Hansen¹, and Sune Nørhøj Jespersen^1,2
1CFIN / MindLab, Aarhus university, Aarhus, Denmark, ²Department of physics and astronomy, Aarhus University, Denmark

Conventionally DTI is deemed accurate for b-values below 1000 s/mm² and DKI for b-values below 2000 s/mm². By analysing DKI and DTI fits on data from a human volunteer and perfusion fixated rat, we show the DKI model is a better fit for a linearly increasing ratio of voxels, independent on tissue type. Up to one-third of voxels are better fitted by DKI when maximum b-value is 1000 s/mm². DKI is a convincing alternative to DTI, even at smaller b-values that has generally been accepted. Also, we show that the fixation process indeed restricts diffusion; this is disputed in the literature.

Ezequiel Farrher¹, Farida Grinberg^1,2, Ivan I. Maximov¹, and N. Jon Shah^1,2
1INM - 4, Research Centre Jülich GmbH, Jülich, Germany, ²Department of Neurology, RWTH Aachen University, Aachen, Germany

Diffusion kurtosis imaging has become an important extension to diffusion tensor imaging in the quantification of non-Gaussian diffusion by means of magnetic resonance imaging. However, there is still no general consensus on the optimal range of b-values for data analysis. In this work we perform an investigation on the dependence of the cumulant expansion terms of the diffusion-weighted MRI signal on the fitting b-value range. The present analysis gives important insights into the validity of the fitting b-value range in DKI analysis.

Timo Roine¹, Ben Jeurissen¹, Wilfried Philips², Alexander Leemans³, and Jan Sijbers^1
1iMinds-Vision Lab, University of Antwerp, Antwerp, Flanders, Belgium, ²Department of Telecommunications and Information Processing, University of Ghent, Ghent, Flanders, Belgium, ³Image Sciences Institute, University Medical Center Utrecht, Utrecht, Netherlands

Diffusion-weighted magnetic resonance imaging is prone to large partial volume effects (PVEs) due to its large voxel size. In this work, we studied the PVEs with non-white matter (WM) tissue on estimation of the fiber orientation distribution function (fODF) with constrained spherical deconvolution (CSD). We found a loss of precision in the detected fiber orientations and an emergence of a high number of false orientations in the fODF. The effects were most significant with gray matter (GM) partial volume. We propose a modification to CSD to further improve the estimation of the fODF especially near the WM-GM interface.

Farida A Grinberg¹, Ezequiel Farrher¹, Ivan I Maximov¹, and N. Jon Shah^1,2
1INM-4, Forschungszentrum Juelich, Juelich, Germany, ²Department of Neurology, RWTH Aachen University, Aachen, Germany

This work provides a comparative analysis of the four non-monoexponential functions in fitting the diffusion response in a) the Monte Carlo diffusion simulations in a set of cylindrical objects, b) synthetic fibre phantoms with parallel and crossing fibres, and c) in vivo human brain tissue. All functions including the diffusion kurtosis, stretched-exponential, the lognormal- and gamma-distribution functions were shown to provide complimentary information and different contrasts in parameter maps.

Rajesh Kumar¹, Santosh K Yadav¹, Jennifer A Ogren², Mary A Woo², Daniel W Kang³, Paul M Macey², Frisca L Yan-Go⁴, and Ronald M Harper^5
1Anesthesiology, University of California at Los Angeles, Los Angeles, CA, United States, ²UCLA School of Nursing, University of California at Los Angeles, Los Angeles, CA, United States, ³Medicine, University of California at Los Angeles, Los Angeles, CA, United States, ⁴Neurology, University of California at Los Angeles, Los Angeles, CA, United States, ⁵Neurobiology, University of California at Los Angeles, Los Angeles, CA, United States

Obstructive sleep apnea (OSA) patients show structural injury and functional deficits in multiple brain areas, based on various MRI techniques, in sites that regulate autonomic, motor, cognitive, and mood functions, deficient in the condition. However, the nature of tissue injury remains unclear. We examined global and regional mean kurtosis values in newly-diagnosed, treatment-naive OSA patients, and found increased global values in critical autonomic, cognitive, motor, and respiratory control sites, including the insular cortices, cerebellum, and basal-ganglia. The pathological mechanisms contributing to tissue injury likely include hypoxemia-induced processes, leading to acute tissue changes in OSA.

Phillip Zhe Sun¹, Yu Wang^1,2, Emiri Mandeville³, Eng H Lo³, and Xunming Ji^2
1Radiology, Martinos Center for Biomedical Imaging, Charlestown, MA, United States, ²Cerebrovascular Diseases Research Institute, Xuanwu Hospital of Capital Medical University, Beijing, China, ³Department of Radiology and Neurology, Neuroprotection Research Laboratory, MA, United States

Diffusion kurtosis MRI (DKI) is a complementary index to stratify heterogeneous DWI deficit. However, DKI acquisition time is relatively long. Here, we tested a fast DKI approach, recently proposed by Hansen et al., using a rodent model of acute stroke. We found significant correlation, per pixel, between the diffusion and kurtosis coefficients measured using the standard and fast DKI protocols. We further demonstrated that diffusion and kurtosis lesion sizes defined using the fast and conventional DKI methods were in good agreement. Moreover, we showed that kurtosis lesion size was significantly smaller than that of diffusion lesions (72±16%, P<0.01, Paired-t test).

Mariko Yoshida¹, Masaaki Hori¹, Kazumasa Yokoyama², Nozomi Hamasaki¹, Michimasa Suzuki¹, Koji Kamagata¹, Kohei Kamiya¹, Shigeki Aoki¹, Yoshitaka Masutani³, and Nobutaka Hattori^2
1Department of Radiology, Juntendo University School of Medicine, Bunkyo, Tokyo, Japan, ²Department of Neurology, Juntendo University School of Medicine, Tokyo, Japan, ³Division of Radiology and Biomedical Engineering, Graduate School of Medicine, The University of Tokyo, Tokyo, Japan

The purpose of this study is to evaluate diffusional changes of the WM and GM including plaques, periplaque WM (PWM), periplaque GM (PGM), NAWM, and NAGM in MS by using a new method�\DKI with double inversion recovery (DIR) to segment GM. The participants were 7 MS patients with GM plaques. DKI with DIR detected abnormalities in WM and GM with high sensitivity and can provide additional information on changes of WM and GM in MS. In contrast to WM plaques, there may be little change in the GM substance around GM plaques. MK may be a more sensitive biomarker of tissue damage in MS patients.

Gene Young Cho¹, Jeff L Zhang², Linda Moy^1,3, Steven Baete¹, Melanie Moccaldi³, Sungheon Kim¹, Daniel K Sodickson¹, and Eric E Sigmund^1
1Radiology - Bernard and Irene Schwartz Center for Biomedical Imaging, New York University School of Medicine, New York, Select, United States,²Radiology, University of Utah, Salt Lake City, UT, United States, ³Radiology, New York University Cancer Institute, New York, NY, United States

The selection of diffusion gradients for diffusion weighted imaging (DWI) is important for protocol optimization. It has been shown that different ranges of gradients sensitize the signal to different components of diffusion. Additionally, quantification of DWI data with the biexponential IVIM fitting model can be numerically nontrivial. In this study, we employ an optimized b value selection for biexponential IVIM analysis of breast lesions to increase measurement precision. Eventually, optimization may lead to increasing diagnostic differentiation between cancerous subtypes and potentially allow more careful monitoring of response to treatment.

Masaaki Hori^1,2, Yoshitaka Masutani³, Ryuji Nojiri², Katsutoshi Murata⁴, Koji Kamagata¹, Mariko Yoshida¹, Michimasa Suzuki¹, and Shigeki Aoki^1
1Radiology, Juntendo University School of Medicine, Tokyo, Japan, ²Tokyo Medical Clinic, Tokyo, Japan, ³The University of Tokyo, Tokyo, Japan, ⁴Siemens Japan K.K., Tokyo, Japan

The purpose of this exhibit is to characterize, particularly for clinical use, the effects of computation methods, median filtering and Rician noise removal on diffusion tensor imaging (DTI) and diffusional kurtosis imaging (DKI) metrics in normal white matter and brain tumors. Median filtering is an important factor that affects quantitative diffusion metrics; for example, it changes fractional anisotropy (FA) values by 10%. Differences of computing methods and Rician noise removal appear to be less influential in changing diffusion metrics. Selection of post-processing methods should be clarified in research and clinical use.

Tomohiro Namimoto¹, Kosuke Morita¹, Yuuki Kizaki¹, Toshinori Hirai¹, Makoto Gotoh², Yukimi Waki², Ryo Itatani¹, and Yasuyuki Yamashita^1
1Diagnostic Radiology, Kumamoto University, Kumamoto, Kumamoto, Japan, ²Kumamoto University Hospital, Kumamoto, Japan

We compared apparent diffusion coefficients (ADCs) with T1-/T2-relaxation time in a phantom study. Blood phantoms were gradually diluted with saline. MR measurements were performed with T1-/T2-calculations and ADCs from 0.2 hour to 7 days. T1-relaxation time cumulatively increased and subsequently plateaued. T2-relaxation time increased, and then decreased. ADC slightly decreased during whole times. T1-relaxation time and ADC was almost proportional decreased as blood concentration decreased; T2-relaxation time was almost inversely proportional decreased. Temporary change of ADCin bloody fluid was poorly correlated with that of T1-/T2-relaxation time. ADC is almost linearly dependent on blood concentration and almost independent of T1-/T2-relaxation times.

Prativa Sahoo¹, Rakesh Kumar Gupta², Indrajit Saha¹, Bhaswati Roy², Ram Krishna Singh Rathore³, and Divya Rathore^4
1Philips Healthcare, Philips India Ltd, Gurgaon, Haryana, India, ²Radiology and Imaging, Fortis Memorial Research Institute, Gurgaon, Haryana, India,³Mathematics and Statistics, Indian Institute of Technology Kanpur, Kanpur, UP, India, ⁴ADISL India, Kanpur, UP, India

Diffusion weighted imaging (DWI) is a very sensitive imaging sequence for detecting various tissue abnormalities in brain. In clinical practice, b-values of 1000 sec/mm2 or less are commonly used; however it has been reported that use of higher b-values improves disease visualization and detection1. DWI imaging with higher b-values is challenging because it suffers with low signal to noise ratio (SNR), distortion along with longer scan time. To overcome these difficulties computed DWI technique has been proposed by many researchers 2. Computed DWI technique is a mathematical technique, which generates images of higher b-values by using at least two different lower b value (b ≤ 1000) images. Here we report that we have observed ADC depends on b-values, on contrary to the previous assumption where it was considered constant for all b-values and computation of ADC corresponding to higher b-values was not possible. The aim of this study was to find the relationship between ADC and b-values for quantification of ADC corresponding to higher b-values using lower b-values with computed DWI technique.

Ivan I. Maximov¹, Anton Minnekhanov^1,2, Ezequiel Farrher¹, Farida Grinberg¹, and N. Jon Shah^1,3
1INM-4, Forschungszentrum Juelich, Juelich, Germany, ²Department of Physics, Moscow State University, Moscow, Russian Federation, ³Department of Neurology, RWTH Aachen University, Aachen, Germany

Diffusion tensor imaging is a well-known and well-established imaging technique. It allows one to easily visualize the neuronal tissue architecture by using the eigenvector linked to the largest eigenvalue of the diffusion tensor. However, the diffusion tensor eigenvectors form a complete orthonormal basis in space and contain additional information. We hypothesize here that all three eigenvectors in the neuronal bundles can generate a structure similar to a crystal lattice and keep this configuration along the bundle with high anisotropy. Thus, vector fields based on secondary and/or tertiary eigenvectors could improve the segmentation/reconstruction of fibre tracts and provide additional information on white matter organisation.

Condon Lau¹, Sherwin Abdoli², Leon C Ho^3,4, Jevin W Zhang^3,4, and Ed X Wu^3,4
1Division of Biomedical Engineering, HKUST, Kowloon, Hong Kong, ²Keck School of Medicine, University of Southern California, CA, United States,³Department of Electrical and Electronic Engineering, The University of Hong Kong, Hong Kong, Hong Kong, ⁴Laboratory of Biomedical Imaging and Signal Processing, The University of Hong Kong, Hong Kong, Hong Kong

Diffusion tensor imaging was used to observe changes in the brain structure of rats (n=16) following two months of continuous and passive acoustic exposure at moderate sound pressure level. Voxel-based statistics (VBS) revealed greater fractional anisotropy of the pyramidal tract, tectospinal tract, intertwined trigeminothalamic tract and medial lemniscus of the exposed rats than of the control rats. ROIs were drawn for the structures indicated and applied to the fractional anisotropy and mean diffusivity maps. ROI analysis confirmed that in the structures indicated by VBS, fractional anisotropy was higher and mean diffusivity was lower in exposed rats than in normal rats.

Atsushi Tachibana¹, Takayuki Obata¹, Yasuhiko Tachibana¹, Hiroshi Kawaguchi², Jeff Kershaw², Ichio Aoki², Hiroshi Ito², and Hiroshi Tsuji^1
1Research Center for Charged Particle Therapy, National Institute of Radiological Sciences, Chiba-shi, Chiba, Japan, ²Molecular Imaging Center, National Institute of Radiological Sciences, Chiba-shi, Chiba, Japan

This study investigates the diffusion characteristics of our newly developed DTI phantom made of bundled polyethylene fibers (Dyneema (Dy)), in comparison to a conventionally used glass made phantom with micro capillaries (capillary plates (CP)). Dy was less sensitive to diffusion time, implying that the water diffusion is hindered rather than restricted. Dy also had b-value dependent signal attenuation more similar to that in vivo tissue, and was much less expensive than CP. These suggest that Dy is a better phantom for DTI QC than CPs.

Devon M. Middleton^1,2, Nadia Barakat³, Sphoorti Shellikeri⁴, Scott H. Faro^1,2, MJ Mulcahey^3,5, and Feroze B. Mohamed^1,2
1Department of Engineering, Temple University, Philadelphia, PA, United States, ²Department of Radiology, Temple University, Philadelphia, PA, United States, ³Shriners Hospital for Children, Philadelphia, PA, United States, ⁴Childrens Hospital of Philadelphia, Philadelphia, PA, United States, ⁵Thomas Jefferson University, Philadelphia, PA, United States

This study examines the effects of three different tensor estimation methods (linear least squares, robust outlier rejection, and linear least squares with non-positive tensor removal) on FA and MD in the cervical spinal cord for pediatric subjects. Both typically developing and injured subjects were examined.

Michael A. Boss¹, Thomas L. Chenevert², John C. Waterton^3,4, David M. Morris⁴, Hossein Ragheb⁴, Alan Jackson⁴, Nandita deSouza⁵, David J. Collins⁵, Bernard E. van Beers⁶, Philippe Garteiser⁶, Sabrina Doblas⁶, Stephen E. Russek¹, Kathryn E. Keenan¹, Edward F. Jackson⁷, and Gudrun Zahlmann^8
1National Institute of Standards and Technology, Boulder, CO, United States, ²University of Michigan, Ann Arbor, MI, United States, ³AstraZeneca, Macclesfield, United Kingdom, ⁴University of Manchester, Manchester, United Kingdom, ⁵The Institute of Cancer Research, Sutton, Surrey, United Kingdom,⁶INSERM, France, ⁷University of Wisconsin, Madison, WI, United States, ⁸F. Hoffman-La Roche Ltd., Basel, Switzerland

We designed and built an ADC phantom to cover a wide range of physiologically relevant ADC values. The phantom used aqueous solutions of polyvinylpyrrolidone (PVP), and we maintained constant temperature by filling the phantom with ice water. The phantom was scanned at three sites (1.5 and 3.0 T) to assess variability of ADC measurement. We found excellent intrascanner repeatability and good reproducibility across systems and field strengths. There was non-negligible variability only for the highest concentrations of PVP, representing ADCs less than 0.3 x 10^-3 mm²/s. The ADC of water at 0 °C was in excellent agreement with literature values of 1.1 x 10^-3 mm²/s.

Dariya Malyarenko¹, David Newitt², Alina Tudorica³, Robert Mulkern⁴, Karl G. Helmer⁵, Michael A. Jacobs⁶, Lori Arlinghaus⁷, Thomas Yankeelov⁷, Fiona Fennessy⁴, Wei Huang³, Nola Hylton², and Thomas L. Chenevert^1
1Radiology, University of Michigan, Ann Arbor, MI, United States, ²Radiology and Biomedical Imaging, University of California San Francisco, San Francisco, CA, United States, ³Oregon Health and Science University, Portland, OR, United States, ⁴Dana Faber Harvard Cancer Center, Harvard Medical School, Boston, MA, United States, ⁵Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital, Boston, MA, United States, ⁶John Hopkins University School of Medicine, Baltimore, MD, United States, ⁷Vanderbilt University Institute of Imaging Science, Vanderbilt Unversity, Nashville, TN, United States

Multi-center clinical trials seek to establish confidence levels for quantitative diffusion measurements. Scanner-specific gradient nonlinearity bias was observed for off-center measurements and implicated as a major source of potential reproducibility error in ADC mapping for some scanners. A practical procedure was developed to empirically characterize the systematic nonlinearity bias across diverse clinical MRI platforms. The results are demonstrated for representative MRI scanners utilized in clinical oncology trials supported by the NCI Quantitative Imaging Network (QIN).

Yen-Peng Liao¹, Shin-ichi Urayama¹, and Hidenao Fukuyama^1
1Human Brain Research Center, Kyoto University, Kyoto, Japan

Small blood volume of brain limits the application of IVIM-MRI. Hypoperfusion is an important index for diagnosis of some brain diseases. However, without sufficient signal-noise ratio, large estimation errors may be yielded in conditions of small cerebral blood volume (CBV) and cerebral blood flow (CBF). In this study, we proposed a method using conventional diffusion weighted imaging combined with flow-compensated pulses to estimate the CBF-related parameter. Computer simulation was performed to evaluate the accuracy and precision in heterogeneous brain tissues. The results showed that the proposed method can estimate extremely low CBF with better precision than the conventional fitting method.

Anna Varentsova¹, Shengwei Zhang², and Konstantinos Arfanakis^2,3
1Department of Physics, Illinois Institute of Technology, Chicago, IL, United States, ²Department of Biomedical Engineering, Illinois Institute of Technology, Chicago, IL, United States, ³Rush Alzheimer's Disease Center, Rush University, Chicago, IL, United States

The IIT Human Brain Atlas project is in the 4th year of development (www.nitrc.org/projects/iit2). The latest version of the atlas contains high-quality artifact-free anatomical, DTI, HARDI templates, and probabilistic gray matter (GM) labels of the adult human brain in ICBM-152 space. The project is currently focusing on the development of a probabilistic white matter atlas that is complementary to the GM atlas, by performing tractography in the state-of-the-art HARDI template using the labels of the GM atlas as seeds. The present work aimed to: a) evaluate the tractography results, and b) generate probabilistic connectivity maps for pairs of GM labels.

Franco Pestilli¹, Justin Solomon², Adrian Butscher³, and Brian Wandell^1
1Psychology, Stanford University, Stanford, CA, United States, ²Computer Science, Stanford University, Stanford, CA, United States, ³Max Planck Institute for Informatics, Saarbruecken, Germany, Germany

We present a method for validating the organization of the white-matter tracts generated using diffusion weighted MRI and fiber tractography. Unlike previous techniques, our method locally determines the quality of each fascicle by associating nodes on each fascicle with nonnegative weights. These weights are optimized to predict the diffusion signal while also exhibiting smoothness along the fiber path. We find that different tractography algorithms generate different white-matter geometries. After cross-validation using our localized fascicle weights, however, we can describe the geometry of fascicles in white matter with confidence that our result is not a by-product of the choice of tractography techniques.

Louise Emsell^1,2, Thijs Dhollander², Koen Nelissen³, Ronald Peters^1,2, Wim Vanduffel³, and Stefan Sunaert^1,2
1Translational MRI, Imaging & Pathology, KU Leuven & Radiology, University Hospitas Leuvenl, Leuven, Belgium, ²Medical Imaging Research Center, Leuven, Belgium, ³Laboratorium voor Neuro-en Psychofysiologie, KU Leuven Medical School, Leuven, Belgium

Visualising white matter anatomy in vivo at submilimetre resolution using diffusion MRI is challenging without access to high-field and strong gradient systems. Short-tracks TDI (stTDI) is a post-processing technique that allows super-resolution visualization beyond the resolution of acquired data. In this pilot study we used stTDI to enhance anatomical contrast in 3T monkey data from 1.0mm3 to 0.25mm3 allowing the identification of structures not visible on the 1.0mm3 colour FA maps (e.g. stria medullaris, pre- and post-commissural fornix, decussation of the superior cerebellar peduncles). Reliability was assessed using bootstrapping to model the effect of noise in the stTDI maps

Ping-Hong Yeh¹, Erick Jorge Canales- Rodríguez², John Morissette³, Arman A. Taheri³, Li-Wei Kuo⁴, Fang-Cheng Yeh⁵, Wei Lui³, John Ollinger³, Terrence Oakes³, Mark L. Ettenhofer⁶, and Gerard Riedy^3
1Henry Jackson Foundation for the Advancement of Military Medicine, Bethesda, MD, United States, ²FIDMAG Research Foundation, Germanes Hospitalaries and CIBERSAM, Barcelona, Spain, ³National Intrepid Center of Excellence, Walter Reed National Military Medical Center, MD, United States, ⁴National Health Research Institutes, Taiwan, ⁵Carnegie Mellon University, PA, United States, ⁶Uniformed Services University of the Health Sciences, MD, United States

A variant of Richardson-Lucy (RL) spherical deconvolution (SD) algorithm modeling Rician noise from high angular resolution diffusion imaging (HARDI) data has been shown efficient in recovering the fiber orientation distribution in the low and medium SNR range. We evaluate the performance of RL_SD reconstruction on undersampled multiple q-shell HARDI data acquired by a clinical 3T MR scanner within an acceptable time frame (< 15 minutes).

Niklas Kasenburg^1,2, Matthew Liptrot¹, Karsten M. Borgwardt^2,3, Silas N. Ørting¹, Mads Nielsen¹, and Aasa Feragen^1,2
1University of Copenhagen, Copenhagen, Zealand, Denmark, ²Max Planck Institutes Tübingen, Tübingen, Baden-Württemberg, Germany, ³Eberhard Karls Universität Tübingen, Baden-Württemberg, Germany

We present a novel, graph-based fibre tractography method, where the graph model can integrate any diffusion representation provided as an ODF. Tractography is performed by computing all possible shortest paths between regions of interest. For each path a confidence value is computed to assess its quality. Voxels are scored by the average value of paths they lie upon. Results for three different known tracts show the robustness of the method over five healthy subjects and a high overlap with tract atlases. The region based approach is especially useful for creating brain connectivity graphs while not suffering from path length dependencies.

Gabriel Girard^1,2, Kevin Whittingstall³, Rachid Deriche², and Maxime Descoteaux^1
1Sherbrooke Connectivity Imaging Lab (SCIL), Computer Science Department, Université de Sherbrooke, Sherbrooke, Qc, Canada, ²ATHENA Project-Team, INRIA, Sophia-Antipolis, France, ³Department of Diagnostic Radiology, Faculty of Medicine and Health Science, Université de Sherbrooke, Sherbrooke, Qc, Canada

Diffusion-weighted imaging is often used as a starting point for in vivo white matter connectivity to reconstruct potential white matter pathways between brain areas. Tractography algorithms have many parameters which influence the reconstruction and connectivity analysis. But how does one choose the best set of parameters? In this study, we varied some of the tractography parameters and observed the connectivity score using the Tractometer evaluation system on the ISBI Challenge 2013 dataset. We think this provides useful information to choose tractography algorithms and its parameters for connectivity study.

Fernando Calamante^1,2, Robert Elton Smith¹, J-Donald Tournier^1,3, David Raffelt¹, and Alan Connelly^1,2
1The Florey Institute of Neuroscience and Mental Health, Melbourne, Victoria, Australia, ²Department of Medicine, Austin Health and Northern Health, University of Melbourne, Melbourne, Victoria, Australia, ³Centre for the Developing Brain, King's College London, London, United Kingdom

A biological parameter that would be valuable to draw from a DWI experiment is the local white matter axonal density. It has been suggested that track-density imaging (TDI) could provide such a measure; however, this has been the subject of controversy. The post-processing method of SIFT was recently introduced to minimise tractography biases. Importantly, the TDI intensity following SIFT ideally corresponds to the DC term of the FOD spherical harmonic expansion. We characterise the TDI reproducibility with and without SIFT pre-processing, as well as the reproducibility of DC maps, to determine the most appropriate method for quantifying voxel-wise fibre densities.

Alia Lemkaddem¹, Didrik Skiöldebrand¹, Alessandro Dal Palú², Jean-Philippe Thiran^1,3, and Alessandro Daducci^1
1Signal Processing Laboratory (LTS5), Ecole Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Switzerland, ²Department of Mathematics and Computer Science, University of Parma, Italy, ³Department of Radiology, University Hospital Center (CHUV) and University of Lausanne (UNIL), Switzerland

Despite the improved robustness of current global tractography methods, the computation time makes them still not clinically feasible. Furthermore, the inclusion of anatomical priors are missing, i.e that the fiber starts and terminates in the gray matter. The latter constraint is thus crucial when performing connectivity studies, as it is mandatory to asses which pair of cortical regions a fiber is connecting. In this work, we have reformulated global tractography to reduce the computation time and furthermore embedded anatomical priors. We compare our results to state-of-the-art global tractography methods and therefore show how we improve.

Birkan Tunc¹, William A. Parker¹, and Ragini Verma^1
1Department of Radiology, University of Pennsylvania, Philadelphia, PA, United States

In this work, we design a framework for extraction of fiber bundles by introducing a group-wise consistent fiber clustering. This consists of defining a connectivity based fiber representation, using this representation to design a fiber atlas, and finally a subsequent adaptive clustering of a new subject based on this atlas. The connectivity based fiber representation encodes information related to the connectivity signatures of fibers. The fiber atlas is generated using several representative scans. The fibers of a new subject is clustered adaptively by incorporating this atlas as a prior. This whole framework results in a population consistent automated fiber clustering.

Jian Lin¹, Ken Sakaie¹, Myron Zhang², Katherine A Koenig¹, Stephen E Jones¹, and Mark J Lowe^1
1Cleveland Clinic, Cleveland, OHIO, United States, ²The Ohio State College of Medicine, OSU, Columbus, OHIO, United States

Tractography can be sensitive to a number of parameters. However, there are few methods for quantifying such sensitivity. We demonstrate a means for quantitative comparison of pathways by comparison with an ensemble of tracks. We apply the method to compare the results of two probabilistic tractography methods: a computationally intensive Monte Carlo (MC) approach and fast partial differential equations (PDE) that is designed to rapidly replicate the results of the MC approach.

Yao-Chia Shih^1,2, Yun-Chin Hsu², Horng-Huei Liu³, and Wen-Yih Isaac Tseng^2
1Institute of Biomedical Engineering, National Taiwan University, Taipei, Taipei, Taiwan, ²Center for Optoelectronic Medicine, National Taiwan University College of Medicine, Taipei, Taipei, Taiwan, ³Department of Neurology, National Taiwan University Hospital, Taipei, Taipei, Taiwan

Patients with unilateral mesial temporal sclerosis(MTS) were shown to have extensive gray matter (GM) and white matter (WM) abnormalities ipsilateral and contralateral to the seizure onset area. The changes of the whole brain WM tracts in MTS still remain unclear. In this study, we proposed a new method to perform tract-specific analysis over the whole brain, named tract-based automatic analysis (TBAA), using a diffusion spectrum imaging (DSI) template and a tract atlas. We hypothesized that left and right MTS would affect WM tract integrity over the whole brain and have different epileptic networks. In conclusion, we successfully identified different epileptic networks in left and right MTS, and found more extensive WM alterations in right MTS than in left MTS.

Jennifer Andreotti¹, Kay Jann^1,2, Lester Melie-Garcia^1,3, Stéphanie Giezendanner¹, Thomas Dierks¹, and Andrea Federspiel^1
1Department of Psychiatric Neurophysiology, University Psychiatric Hospital, Bern, BE, Switzerland, ²Department of Neurology, University of California Los Angeles, Los angeles, California, United States, ³Department of Neuroinformatics, Cuban Neuroscience Center, Havana, Havana, Cuba

Communicability is a wider measure of network connectivity based on the idea that any path between two nodes will contribute to the total flow of information. Previous studies suggest that communicability may be sensitive to reorganizational changes of the brain network following a lesion. In our study the sensitivity of communicability related metrics was assessed using simulated lesions modelled as attacks to nodes and single edges. Our analysis confirmed that communicability metrics are an interesting tool to study the effects of lesions as they are sensitive to changes in the networks also in regions distant from the main lesion focus.

Namita Singh Saini¹, Giriraj Singh Gujral², Richa Trivedi³, Archana Kumari³, Prabhjot Kaur³, Pawan Kumar³, Subhash Khushu³, and Rajender Prasad Tripathi^3
1NMR, INMAS, Delhi, Delhi, India, ²Radiodiagnosis, Command Hospital Southern Command, Pune, Maharashtra, India, ³INMAS, Delhi, India

Neurological damage in HIV infection ranges from Minor Cognitive Motor Dysfunction in the neuroasymptomatic stage to frank HIV associated dementia later, necessitating placement under Highly Active Anti Retroviral Therapy (HAART). Conventional MRI is unable to detect abnormalities in the early stages while Diffusion Tensor Imaging shows reduced Fractional Anisotropy (FA) in Corpus Callosum in later stages. In a cohort of neuroasymptomatic HIV positive individuals we have found evidence of motor progression and reduced FA in Corticospinal Tract even in subjects with CD4 counts above 250, indicating disease progression. Thus DTI instead of CD4 could determine necessity for HAART cover.

Jeong-Won Jeong^1,2
1Pediatrics and Neurology, Wayne State University, Detroit, Michigan, United States, ²PET center, Children's Hospital of Michigan, Detroit, Michigan, United States

The present study investigates whether independent component analysis with ball-stick model (ICA+BSM) analysis can estimate accurate orientation of multiple fiber bundles in clinical DTI. Simultaion studies assessing the absolute errors for two and three fibers per voxel demonsrated that ICA+BSM provides promising accuracy (median value of error angle = 4.1 ° and 10.4 ° for two and three fiber crossing, respectively) even at b-value =1000 s/mm² and 12 encoding gradients. A preliminary human study suggests these errors may be marginal to isolate three crossing pathways, copus callosum, arcuate fasciculus, and cortico-spinal tracts in the lateral regions of central gyrus/sulcus.

Durai Arasan¹, Wang Zhan¹, and Luiz Pessoa^1
1University of Maryland, College Park, MD, United States

Compared with diffusion tensor imaging (DTI) techniques, diffusion spectrum imaging (DSI) has demonstrated its superiority in characterizing complex diffusion patterns related to multiple fiber structures within a single imaging voxel. However, the test-retest reliability of DSI-based tractography remains unclear since assessments so far were generally based on DTI only. Here we present a study to investigate the test-retest reliability of a set of tracts in a group of healthy subjects, with 4 independent DSI scans for each. The outcome tractography measurements, such as fiber counts and tract length, show higher test-retest reliability than previously reported DTI results.

Jennifer Andreotti¹, Alessandra Griffa², Thomas Dierks¹, Andrea Federspiel¹, and Patric Hagmann^2,3
1University Psychiatric Hospital, Bern, BE, Switzerland, ²Swiss Federal Institute of Technology, Lausanne, VD, Switzerland, ³CHUV, Lausanne, VD, Switzerland

Quantitative tractography methods require a detailed characterization of tissue properties specific to each white matter (WM) fiber tract. In our study Monte-Carlo simulations are used to understand how the generalized fractional anisotropy (GFA) is affected in the case of variable configurations of WM fiber tracts crossings. In particular, a linear sum of specific anisotropy measures for each of the tracts (linGFA) is compared to the usual GFA. The analysis shows that, compared to standard GFA, linGFA has a more linear relationship with the underlying characteristics of each of the fiber tracts and hence would be beneficial for quantitative tractography studies.

Ek T Tan¹, Xiaofeng Liu¹, Aziz M Ulug², Peter B Kingsley³, Anil K Malhotra^2,3, Delbert G Robinson^2,3, Philip R Szeszko^2,3, and Luca Marinelli^1
1GE Global Research, Niskayuna, NY, United States, ²Feinstein Institute for Medical Research, Manhasset, NY, United States, ³North Shore-LIJ Health System, Manhasset, NY, United States

Brain tractography is a manually-intensive and operator-dependent task, requires anatomical knowledge, and is inaccurate in regions containing crossing-fibers. We propose a robust, automated seeding method that performs image registration (1) to a predefined seeding template, and (2) multi-fiber tractography using compressed-sensing-accelerated diffusion spectrum imaging (CS-DSI). Because the method uses simple spherical ROIs, seeds can be easily adjusted on a manual basis to optimize tract identification. The automated algorithm was evaluated on CS-DSI data acquired in six healthy volunteers and seven patients with schizophrenia. The proposed method simplifies tractography workflow, and may provide improved reproducibility in studies utilizing diffusion tractography.

Rafael Neto Henriques^1,2, Luis Lacerda^2,3, Rita Nunes², Marta Morgado Correia¹, and Hugo Alexandre Ferreira^2
1Cognition and Brain Science Unit, Medical Research Council, Cambridge, United Kingdom, ²Institute of Biophysics and Biomedical Engineering, Faculty of Sciences of the University of Lisbon, Lisbon, Portugal, ³Centre for Neuroimaging Sciences, Institute of Psychiatry, King's College London, London, United Kingdom

Diffusion kurtosis Imaging (DKI) is an extension of diffusion tensor imaging. In addition to the diffusion tensor, DKI estimates a diffusion kurtosis tensor which provides an indirect measure of tissue heterogeneity. Moreover, previous studies show that the diffusion kurtosis tensor can help resolve crossing fibres. Thus, DKI-based tractography algorithms can be advantageous for visualization of crossing white matter pathways in clinical applications since they require less data and hence shorter scanning times relative to HARDI methods. Here, two tractography algorithms based on DKI are evaluated, one of which is shown to be able to resolve fibers crossing up to 45º.

Caterina Gaudiano¹, Fiorenza Busato¹, Beniamino Corcioni¹, Valeria Clementi², Emma Fabbri³, Paola Berardi³, and Rita Golfieri^1
1Radiology Unit, Sant'Orsola-Malpighi Hospital, Bologna, Bologna, Italy, ²Clinical Science Development Group, GE Healthcare, Bologna, Italy, ³Medical Physics Department, Sant'Orsola-Malpighi Hospital, Bologna, Bologna, Italy

We studied 17 renal patients with RAS detected by using CE-MRA and 15 control subjects with 1.5 T MRI scanner adding DTI to the standard protocol. The comparison between the two groups showed that medullary ADC/FA in the study group were significantly lower than in the control group (P=0.034 and 0.022, respectively); there was no significant difference for the other parameters. This suggests that DTI could be a useful tool in evaluating renal alteration in hemodynamic RAS.

Jonathan Ashmore¹, Flavio Dell'Acqua², Ruth O'Gorman³, Gareth Barker², and Jozef Jarosz^1
1Neuroradiology, Kings College Hospital, London, United Kingdom, ²Department of Neuroimaging, Kings College London, London, United Kingdom, ³University Children's Hospital, Zurich, Switzerland

The purpose of this study was to investigate how diffusion acquisition parameters including the number of diffusion directions, the use of cardiac gating and the use of non-isotropic versus isotropic voxels impacts on tractography of the corticospinal tract for application to neurosurgical planning. We consider the tractography generated from a regulatory approved neurosurgical planning system to those produced from research software which incorporates constrained spherical deconvolution and probabilistic tractography. All tractography results are assessed against cyctoarchitectonic maps of the corticospinal tract which is used as a gold standard to asses the validity of the generated tracts.

Stefan Sommer^1,2, Sebastian Kozerke¹, Erich Seifritz², and Philipp Staempfli^2
1Institute for Biomedical Engineering, ETH Zurich, Zurich, Zurich, Switzerland, ²Departement of Psychiatry, Psychotherapy and Psychosomatics, Zurich University Hospital of Psychiatry, Zurich, Zurich, Switzerland

A fundamental problem of tracking techniques based on high angular resolution diffusion data is that a quantitative evaluation of the resulting fiber tracts is still error prone and dependent on the algorithm and anatomy of interest. Furthermore, it remains challenging to derive e.g. crossing ratios of two or more intersecting fiber populations in a robust and reproducible manner. In this work, we present a method to minimize fiber tracking errors of any tracking algorithm in order to improve the quality of fiber tracts and fiber quantification of fiber crossing ratios. The method is validated using synthetic data and in-vivo data.

Chantal M.W. Tax¹, Marijn van Stralen¹, Max A. Viergever¹, Nick Ramsey^2,3, and Alexander Leemans^1
1Image Sciences Institute, University Medical Center Utrecht, Utrecht, Utrecht, Netherlands, ²Rudolf Magnus Institute of Neuroscience, University Medical Center Utrecht, Utrecht, Utrecht, Netherlands, ³Department of Neurology and Neurosurgery, University Medical Center Utrecht, Utrecht, Utrecht, Netherlands

The complex nature of diffusion MRI data has triggered the development of new approaches to visualize the architectural organization of diffusion MRI trajectories, which is becoming increasingly important with emerging human connectome studies and the growing interest of applying diffusion MRI tractography in clinical applications. We propose a visualization approach that interactively and selectively visualizes these tracts based on their local orientation, by applying an orientation-dependent transparency rendering to the fiber pathways. This greatly improves the visualization and exploration of the 3D architectural organization of pathways and the underlying tissue configurations that otherwise would be largely covered by other pathways.

Anastasia Yendiki¹, Martin Reuter¹, Paul Wilkens¹, Herminia Diana Rosas¹, and Bruce Fischl^1,2
1HMS/MGH/MIT Athinoula A. Martinos Center for Biomedical Imaging, Charlestown, MA, United States, ²MIT Computer Science and Artificial Intelligence Laboratory, Cambridge, MA, United States

We propose a probabilistic method for reconstructing white-matter pathways from longitudinal diffusion MRI data. We model the posterior probability of a pathway given a subject's full longitudinal data set, including diffusion and structural images from all time points. Our method is unbiased, making no assumptions on the direction of longitudinal change. By design, it allows longitudinal analysis of anisotropy and diffusivity measures to be performed as a function of position along the trajectory of a tract. We demonstrate that our longitudinal tractography improves both specificity and sensitivity compared to the conventional approach of performing tractography in each time point independently.

Stelios Angeli¹, Nicolas Befera², Jean-Marc Peyrat³, Evan Calabrese², George Allan Johnson², and Christakis Constantinides^1,4
1University of Cyprus, Nicosia, Nicosia, Cyprus, ²Duke University Medical Center, Durham, NC, United States, ³Qatar Robotic Surgery Centre, Doha, Qatar,⁴Chi Biomedical Ltd, Nicosia, Cyprus

 

Yu-Chun Lo¹, Yung-Chin Hsu¹, Yu-Jen Chen¹, and Wen-Yih Isaac Tseng^1,2
1Center for Optoelectronic Medicine, National Taiwan University College of Medicine, Taipei, Taiwan, ²Graduate Institute of Brain and Mind Sciences, National Taiwan University, Taipei, Taiwan

White matter tract atlas has gained interests in the neuroscience community because it provides 3D anatomy of white matter fiber pathways in the human brain. Recently, we constructed a diffusion spectrum imaging (DSI) template (NTU-DSI-122) from 122 healthy adults¡¦ DSI datasets. ¡§Difficult¡¨ fiber pathways which are of small sizes or have frequent crossings with other fiber tracts were reconstructed on the template. The tract atlas can be used to understand the geometric features of the white matter fiber pathways. Moreover, It allows us to implement a template-based approach that enables the microstructural integrity of the tracts in the human brain.

Nan-Jie Gong^1,2, Chun-Sing Wong¹, Chun-Chung Chan³, Lam-Ming Leung⁴, Yiu-Ching Chu⁵, and Queenie Chan^6
1Diagnostic Radiology, The University of Hong Kong, Hong Kong, China, ²Radiology, Stanford University, Stanford, CA, United States, ³Geriatrics & Medicine, United Christian Hospital, Hong Kong, China, ⁴Psychiatry, United Christian Hospital, Hong Kong, China, ⁵Radiology, Kwong Wah Hospital, Hong Kong, China, ⁶Philips Healthcare, Hong Kong, China

We utilized DKI, and for the first time a white-matter-model that provided metrics of explicit neurobiological interpretations in cognitive aging adults. Whole brain TBSS and regional results suggested that age-related white matter degenerations were broadly driven by axonal loss across nearly all tracts. In the anterior brain which is mostly composed of the late-myelinated fibre tracts, demyelination was also a major mechanism contributing to disintegration. Such probable coexistence of both mechanisms was in line with the Wallerian degeneration theory and is more supportive of the anterior-posterior gradient degeneration than the retrogenesis theory.

Bhaswati Roy¹, Rakesh K Gupta¹, Richa Trivedi², Alok Kumar Singh³, Ravindra Kumar Garg³, Yogita Rai⁴, Vimla Venkatesh⁵, Mukesh Tripathi⁶, and Chandra M Pandey^7
1Radiology & Imaging, Fortis Memorial Research Institute, Gurgaon, Haryana, India, ²NMR Reseach Centre, Institute of Nuclear Medicine and Allied Sciences, Delhi, Delhi, India, ³Neurology, King George's Medical University, Lucknow, Uttar Pradesh, India, ⁴Radiodiagnosis, Sanjay Gandhi Post Graduate Institute of Medical Sciences, Lucknow, Uttar Pradesh, India, ⁵Microbiology, King George's Medical University, Lucknow, Uttar Pradesh, India,⁶Anaesthesiology, Sanjay Gandhi Post Graduate Institute of Medical Sciences, Lucknow, Uttar Pradesh, India, ⁷Biostatistics & Health Informatics, Sanjay Gandhi Post Graduate Institute of Medical Sciences, Lucknow, Uttar Pradesh, India

With an aim to assess the integrity of white matter in HIV positive pediatric patients, DTI and neuropsychological test battery was performed on 38 HIV-seropositive and 8 HIV-seronegative children born to HIV-seropositive mothers of same socioeconomic status. Voxel wise-based technique was implemented for the analysis of DTI data of participants. We observed significant changes in FA values in the corpus callosum, corona radiata, internal capsule. Significant direct correlation was also observed between FA and cognition scores relating to memory function. DT MRI could be used as an early objective marker of HIV related microstructural brain changes.

Mariko Yoshida¹, Masaaki Hori¹, Kazumasa Yokoyama², Nozomi Hamasaki¹, Michimasa Suzuki¹, Koji Kamagata¹, Kohei Kamiya¹, Shigeki Aoki¹, Yoshitaka Masutani³, and Nobutaka Hattori^2
1Department of Radiology, Juntendo University School of Medicine, Bunkyo, Tokyo, Japan, ²Department of Neurology, Juntendo University School of Medicine, Tokyo, Japan, ³Division of Radiology and Biomedical Engineering, Graduate School of Medicine, The University of Tokyo, Tokyo, Japan

The purpose of this study is to evaluate differences in diffusional changes of the optic pathways between multiple sclerosis (MS) and neuromyelitis optica (NMO) using a new method, diffusional kurtosis imaging (DKI) Six NMO, eight MS, and seven healthy volunteers participated in this study. DKI sensitively detected abnormalities in optic tract in NMO patients. According to the past reports, NMO-immunoglobulin G targeting AQP4 is often observed in optic tact. DKI may be a more sensitive biomarker to differentiate between MS and NMO than conventional diffusional evaluations, such as diffusion tensor imaging.

Farida A. Grinberg¹, Ezequiel Farrher¹, Luisa Ciobanu², Françoise Geffroy², Denis Le Bihan², and N. Jon Shah^1,3
1Forschungszentrum Juelich, Juelich, Germany, ²Neurospin, CEA, France, ³RWTH Aachen University, Germany

We compare a sensitivity of two non-Gaussian diffusion models with the gold standard ADC approach in providing the contrast for ischemic lesions in animals: the stretched-exponential and gamma-distribution functions. Both functions provide good fits in the extended range of the diffusion weightings and allow for an enhanced contrast of the lesions. An interesting finding of this work is the appearance of laminar cortical structures in stroke lesions clearly visualised by the parameter maps of the investigated models but hardly observable in the ADC-maps.

Yonggang Lu¹, Vaios Hatzoglou¹, Hilda E. Stambuk¹, Suchandrima Banerjee², Ajit Shankaranarayanan², Yousef Mazaheri¹, Joseph O. Deasy¹, and Amita Shukla-Dave^1
1Memorial Sloan-Kettering Cancer Center, NEW YORK, New York, United States, ²Global Applied Sciences Laboratory, GE Healthcare, Menlo Park, California, United States

This study investigated the repeatability and assessed the radiologic quality of reduced field of view diffusion weighted imaging (rFOV DWI) in thyroid glands of healthy human volunteers. Repetitive DWI scans were performed on each subject within the same scan exam and across scan exams. Image qualities were scored and apparent diffusion coefficient (ADC) values were calculated on the lobes of thyroid glands and followed by reproducibility investigation. The study demonstrated that rFOV DWI had significantly improved image quality but lower ADC values and lesser repeatability value of ADC compared to the conventional DWI techniques in the region of thyroid glands.

Subash Khushu¹, Sadhana Singh¹, Richa Trivedi¹, Kavita Singh¹, Pawan Kumar¹, and L Ravi Shankar^2
1NMR Research Centre, INMAS DRDO, Delhi, delhi, India, ²Thyroid Research Centre, INMAS DRDO, Delhi, delhi, India

The aim of our study was to correlate DTI measures (FA and MD) with memory dysfunction scores in hypothyroid population. Diffusion tensor tractography (DTT) and PGI-memory scale (PGIMS) were performed on eight healthy controls and eight hypothyroid patients. Our results showed the inverse correlation of memory dysfunction score with FA of WM tracts (RIFO and LIFO) and positive correlation of memory dysfunction scores with MD of RATR fibre tract. These findings suggest that the microstructural changes in these WM tracts may contribute to underlying dysfunction in memory in hypothyroidism.

Oi Lei Wong^1,2 and Michael D Noseworthy^3,4
1Department of Medical Physics and Applied Radiation Science, McMaster University, Hamilton, Ontario, Canada, ²Medical Physics and Research Department, Hong Kong Sanatorium & Hospital, Hong Kong, Hong Kong, Hong Kong, ³Department of Electrical and Computer Engineering, McMaster University, Hamilton, Ontario, Canada, ⁴Imaging Research Center, St.Joseph's Healthcare, Hamilton, Ontario, Canada

Both IVIM and DTI are well recognized MRI methods for quantifying flow and diffusion, respectively. Simultaneously assessing both techniques (IVIM-DTI) has recently been shown feasible in the kidney. We proposed that the same technique is also feasible in the liver with multiple breath holds and subsequent retrospective motion compensation. Based on our result, IVIM-DTI analysis is able to differentiate between hepatic blood vessels and liver parenchyma, while minimizing the pseudo-hepatic artifact.

LI Qiu-ju¹, ZHANG Zi-heng², LI Jia-hui¹, YU Bing¹, ZHANG Xin¹, ZHAO Zhou-she², SHI Yu¹, XIN Jun¹, and GUO Qi-yong^1
1Radiology, Shengjing Hospital of China Medical University, Shenyang, Liaoning, China, ²General Electronic Company Healthcare (China), Beijing, China

A multi-b diffusion weighted MR imaging measurement was performed to examine the effects of AQP inhibition to normal and liver-fibrotic rats in vivo. At low-b values, the resulted ADCs reflected the hepatic blood perfusion information while the functional changes of the AQPs on cell membranes at high b values. The AQP1 expression in the liver endothelial cells increased with aggravation of liver fibrosis. An increase of perfusion was observes at S1, but a decrease occurred with the addition of inhibitor. At S2, perfusion decreased, and further decreased with inhibitor. The inhibition tests were all negative.

Gina Belmonte¹, Vito Biondo², Augusto La Penna³, Francesco Carbone⁴, Gianluca Vischi⁵, Leonardo Semeraro⁵, and Fabrizio Banci Buonamici^1
1Medical Physics, University Hospital of Siena, Siena, Italy, Italy, ²University of Siena, Italy, Italy, ³Radiology, University of Siena, Italy, Italy,⁴Radiology, University Hospital of Siena, Italy, Italy, ⁵Radioterapy, University of Siena, Italy, Italy

We analyzed precision and accuracy in the discrimination of prostate tumor of IVIM parameters measured with a bi-exponential fit and a simplified approach, that allows to estimate the diffusion coefficient and the pseudo-diffusion fraction from a 2 b-values acquisition. ROC analysis were used to extract the optimum cut-off for each parameter that maximize the accuracy of test. The simplified IVIM parameters are not accurate estimates of the corresponding ones from the bi-exponential fit, but discriminate prostate tumor from healthy tissues in a accurate way. This simplified technique allows therefore to reduce the acquisition time and preserve a good diagnostic accuracy.

tao gong^1,2, bin wang¹, lili li², and guangbin wang^2
1binzhou medical university, shandong, yantai, China, ²Shandong Medical Imaging Research Institute, shandong, jinan, China

The DTI parameters of MD/FA values of normal prostate peripheral zone is different in different age groups, and MD values were increased with age , while FA values decreased.

Nan-Han Lu^1,2, Tai-Been Chen², Yung-Hui Huang², Hueisch-Jy Ding², Lee-Ren Yeh¹, and Jeon-Hor Chen^1,3
1Department of Radiology, E-DA Hospital, Kaohsiung, Kaohsiung, Taiwan, ²Department of Medical Imaging and Radiological Sciences, I-Shou University, Kaohsiung, Kaohsiung, Taiwan, ³Functional Onco-Imaging, Department of Radiological Sciences, University of California Irvine, Irvin, CA, United States

Lung cancer is the leading cause of cancer death among men and women worldwide. The precise pre-operation diagnosis of chest neoplasms and inflammatory nodules using CT, MRI and PET-CT is difficult. The most sensitive diagnostic method is the surgical biopsy. The differences in imaging findings may reflect differences in histopathologic features of chest lesions. On T2-weighted MR images could not be used to distinguish malignant from benign pulmonary nodules. On diffusion-weighted MR imaging (DWI), the apparent diffusion coefficient (ADC) value refers to the specific diffusion capacity of a biologic tissue. ADC value depends largely on the presence of barriers to diffusion within the water microenvironment.The goal of this study is to present a novel method for differential diagnosis among primary chest cancers, metastatic tumors and benign tumors using ADC standard deviation value (SD) of diffusion-weighted MR imaging.

Crombé Amandine^1,2, Menegon Patrice¹, Tourdias Thomas¹, and Dousset Vincent^1
1Department of neuroradiology, Pellegrin Hospital, Bordeaux Segalen University, Bordeaux, France, ²ENS Lyon, Lyon, France

From a commercially available axial SE-EPI DTI sequence, after empirical optimizations in order to limit distortion in an acceptable scan time, our aim was to determine the normal values of DTI parameters (mean diffusivity, axial diffusivity, radial diffusivity and fractional anisotropy) in key locations of cervical spine (cortico spinal tract, posterior column and anterior horn, full spine section) at each cervical level from C1 to C7.

Conrad Rockel^1,2 and Michael D Noseworthy^1,2
1School of Biomedical Engineering, McMaster University, Hamilton, Ontario, Canada, ²Imaging Research Centre, St Joseph's Healthcare, Hamilton, Ontario, Canada

Spatial distribution of eigenvector variability was investigated in the human calf at 3T using different amoungs of signal averages and encoding directions. The principal vector was found to vary much less than the two minor vectors. Furthermore, the spatial distribution appeared non-random and favourable to the peripheral musles, although noisier for the minor eigenvectors. The spatial distribution appeared dependent upon scan acquisition conditions.

Yanjie Xue¹, Usha Sinha¹, Vadim Malis¹, Robert Csapo², and Shantanu Sinha^2
1Physics, SDSU, San Diego, CA, United States, ²Radiology, UCSD, San Diego, CA, United States

Age related changes in muscle alter fiber architecture (fiber lengths and pennation angles). Diffusion tensor imaging (DTI) allows the mapping of fiber architecture, this study investigates age related changes in fiber architecture of the medial and lateral gastrocnemius using DTI. Ten female subjects (5 young/ 5 old) were imaged using DTI and fiber length and pennation angles were determined. Decreases in fiber length and pennation angle were seen in the older cohort as well as regional differences between distal and middle origin fibers. The length/pennation angle changes in the LG were smaller than those seen in the MG.

Giulio Gambarota¹, Antoine Marchand², Eric Hitti¹, Frederik Monge¹, Regis Duvauferrier², Raphael Guillin², and Hervé Saint-Jalmes^1
1Université de Rennes 1, LTSI, Rennes, F-35000, France, ²Department of Imaging, Rennes University Hospital, Rennes, F-35000, France

The aim of this study was to assess the feasibility of measuring diffusion and perfusion in vertebral bone marrow using the intravoxel incoherent motion (IVIM) approach and to investigate the benefits of the non-negative least square (NNLS) technique for analysis of IVIM data. Results indicated that, for a proper analysis of IVIM data, the use of NNLS is essential especially in cases where the signal-to-noise ratio is low or where motion artifacts and other factors might affect the data quality.

Yonglan He¹, Huadan Xue², Ning Ding², Yuan Li³, and Zhengyu Jin^4
1Radiology, Peking Union Medical College Hospital, Beijing, Beijing, China, ²Radiology, Peking Union Medical College Hospital, Beijing, China, ³gynaecology and obstetrics, Peking Union Medical College Hospital, Beijing, China, ⁴Peking Union Medical College Hospital, Beijing, China

Prospectively investigate the cyclic changes of FA and ADC values of the normal uterus in a larger population divided into different age groups during 4 phases of the menstrual cycle, and the correlation with the basic serum hormone levels in menstrual phase.

David Newitt¹, Ek Tsoon Tan², Lisa Wilmes¹, Hee Jung Shin³, Luca Marinelli², and Nola Hylton^1
1Radiology and Biomedical Imaging, UCSF, San Francisco, CA, United States, ²GE Global Research, Niskayuna, NY, United States, ³Radiology, Asan Medical Center, University of Ulsan, Korea

Gradient nonlinearity (GN) is a significant source of error for quantitative diffusion MRI and is of particular concern in breast imaging where anatomy requires large offsets from magnet isocenter. GN correction (GNC) to correct for spatially-varying diffusion encoding was evaluated in DTI acquired ADC and FA of tumor and normal tissue regions in breast cancer subjects undergoing neoadjuvant chemotherapy. GNC resulted in a mean change in tumor ADC of -11% (range -4% to -19%), and in FA of -6% (range -24% to 20%). Percent change in tumor ADC between visits had an RMS change of 3% with GNC.

Matthan W.A. Caan¹, Michiel B. de Ruiter¹, Liesbeth Reneman¹, and Aart J. Nederveen^1
1Radiology, Academic Medical Center, Amsterdam, Netherlands

The goal of this study was to compare diffusion parameters between three scanners of one vendor with the head-coils used in daily practice on the particular scanners. Of seven subjects, diffusion weighted MRI data were acquired on 3.0 Tesla Philips Intera, Achieva and Ingenia scanners. Signal to Noise Ratio (SNR), Mean Diffusivity (MD) and Fractional Anisotropy (FA) maps were compared. Inter-scanner differences were three times as large as intra-scanner differences. SNR varied up to 20% between scanners, was positively associated with MD and negatively with FA. Differences in MD and FA up to 7% are in agreement with literature.

Marina Takatsuji¹, Toshiaki Miyati¹, Naoki Ohno¹, Saori Yoshizawa², Tomohiro Noda¹, Satoshi Kobayashi³, Toshifumi Gabata³, and Osamu Matsui^3
1Division of Health Sciences, Institute of Sciences, Kanazawa University, Kanazawa, Ishikawa, Japan, ²School of Health Sciences, College of Medical, Pharmeceutical and Health Sciences, Kanazawa University, Kanazawa, Ishikawa, Japan, ³Department of Radiology, School of Medical Sciences, Kanazawa University, Kanazawa, Ishikawa, Japan

The change of the apparent diffusion coefficient ( ADC)during the cardiac cycle in the brain are affected by the regional cerebral blood flow (rCBF). To evaluate hemodynamic-independent water fluctuaion, we corrected the rCBF effect by using the ECG-triggered single shot diffusion echo planner data itself. We corrected the the changes of ADC (ĢADC) by the maximum ADC with b=0-200 having the highest correlation with the rCBF. There was no significant correlation between corrected-ĢADC and eCBF. Corrected-ĢADC makes it possible to obtain the degree of fluctuation of the water molecules hemodynamic-in the brain.

Evangelia Kaza¹, David Collins¹, Richard Symonds-Tayler¹, Rafal Panek¹, and Martin Osmund Leach^1
1CR-UK and EPSRC Cancer Imaging Centre, Institute of Cancer Research and Royal Marsden Hospital, Sutton, Surrey, United Kingdom

The feasibility of diffusion-weighted (DW) MRI automatically acquired during breath holding enforced by a modified, MR-compatible Active Breathing Coordinator (ABC) system was demonstrated on healthy volunteers. The position of abdominal organs, major vessels and the diaphragm was shown to be reproducible between different measurements under ABC control at the same inhaled air volume and unaffected by changes of non-spatial sequence parameters. The observed repeatability of organ position using the ABC device suggests that DW-MRI with controlled lung volume can be matched to CT images acquired with the same ABC settings and thus support radiotherapy treatment planning.

Julie Hamaide¹, Geert De Groof², Johan Van Audekerke¹, Lisbeth Van Ruijssevelt¹, Zhenhua Mai¹, Firat Kara¹, Marleen Verhoye¹, and Annemie Van Der Linden^1
1University of Antwerp, Wilrijk, Antwerp, Belgium, ²University of Antwerp, Antwerp, Belgium

The aim of the present study was to implement a non-invasive in vivo diffusion tensor imaging (DTI) protocol in the tiny zebra finch. To test for the sensitivity of the optimized sequence, we tried to visualize known (histology) sexual dimorphisms in adult zebra finches. We found the known dimorphisms and illustrate the benefit of using whole-brain in vivo MRI by revealing novel, unexplored sexual dimorphisms in the adult zebra finch brain.

Philip A Cook¹, Paul A Yushkevich¹, Sijie Tian², Sandra Luz³, Seema Bhatnagar⁴, and James C Gee^1
1Department of Radiology, University of Pennsylvania, Philadelphia, PA, United States, ²School of Engineering and Applied Science, University of Pennsylvania, Philadelphia, PA, United States, ³Children's Hospital of Philadelphia, PA, United States, ⁴Anesthesiology and Critical Care, Children's Hospital of Philadelphia and University of Pennsylvania, Philadelphia, PA, United States

TSA in rat brain

Ritu Tyagi¹, Poonam Rana¹, Richa Trivedi¹, B.S. Hemanth Kumar¹, Deepak Bhatnagar², and Subash Khushu^1
1NMR Research centre, Institute of Nuclear Medicine and Allied Sciences (INMAS), DELHI, DELHI, India, ²School of Biochemistry, Devi Ahilya Vishwavidayalaya, Indore, M.P., India

HMTAs are composed of a mixture of tungsten (91–93%), nickel (3–5%) and either cobalt (2–4%) or iron (2–4%) particles which are increasingly adopted as the raw material to make parts of military products. Diffusion tensor imaging (DTI) study was performed at day 1, day 3, and day 5 in controls and animals treated metals used in HMTAs and their mixture. An increase in MD values were observed in cerebral pedeuncle with cobalt and tungsten, whereas all the three metal salts and their mixture showed pronounced effect on the thalamic region of the brain compared to controls.