Emmanuel Caruyer¹, Iman Aganj², Christophe Lenglet³, Guillermo Sapiro², and Rachid Deriche^1
1Athena Project-Team, INRIA Sophia Antipolis - Méditerranée, Sophia Antipolis, France, ²Department of Electrical and Computer Engineering, University of Minnesota, Minneapolis, MN, United States, ³Department of Radiology - CMRR, University of Minnesota Medical School, Minneapolis, MN, United States

We demonstrate how Orientation Distribution Functions (ODFs) can be estimated online from High Angular Resolution Diffusion Imaging (HARDI) data and how this procedure can efficiently be used to detect head motion during acquisition.

Qiuyun Fan^1,2, Xin Hong², Nicole Davis^3,4, Laurie E. Cutting^3,5, and Adam W. Anderson^1,2
1Department of Biomedical Engineering, Vanderbilt University, Nashville, TN, United States, ²Vanderbilt University Institute of Imaging Science, Nashville, TN, United States, ³Vanderbilt University Kennedy Center for Research on Human Development, Nashville, TN, United States, ⁴Department of Radiology and Radiological Sciences, Vanderbilt University, Nashville, TN, United States, ⁵Department of Special Education, Vanderbilt Peabody, Nashville, TN, United States

DTI provides reproducible measures of fiber integrity, although it is unable to resolve crossing fibers. HARDI methods can resolve crossings, but do not provide estimates of the intrinsic anisotropy in each fiber. In this work, we propose an approach that resolves crossing fibers and estimates the intrinsic diffusion properties of each fiber. Multiple kernels are allowed in spherical deconvolution and multiple shells in q-space are sampled in order to estimate the kernel for each resolvable fiber in a voxel. The results can potentially provide more accurate analysis of the properties of fiber pathways in the brain.

Indika S Walimuni¹, and Khader M Hasan^1
1Radiology, UTHSCH, Houston, Texas, United States

A comprehensive investigation of the interplay between microstructural MRI parameters such as T2 relaxation and diffusivites using standardized volume-based methods in both healthy gray and white matter has not been attempted before. Advances in multimodal MRI registration and segmentation methods have enabled the fusion of high resolution T1, diffusion tensor imaging and relaxation maps to a common native space provided by a high resolution T1-weighted volume. This volume can be anatomically labeled using atlases of white and gray matter regionn. In this work, we investigated the interplay between T2 relaxation and the radial diffusivity using 82 brain white matter (WM), gray matter (GM), cortical and subcortical structures. We report, a strong relation between the T2 relaxation and the radial diffusivity in a healthy population.

Alvina Goh¹, Neda Jahanshad², Paul M Thompson², and Christophe Lenglet^3
1Department of Mathematics, National University of Singapore, Singapore, Singapore, ²Laboratory of Neuro Imaging, Department of Neurology, UCLA, Los Angeles, CA, United States, ³Department of Radiology - CMRR, University of Minnesota Medical School, Minneapolis, MN, United States

Orientation Distribution Functions (ODFs) are estimated from High Angular Resolution Diffusion Imaging (HARDI) data and provide a great amount of information on the structure of the white matter. We present an extension of voxel-based morphometry to ODFs and apply it to brain asymmetry.

Lin-Ching Chang¹, and Mikhail A Gorbachev^1
1Department of Electrical Engineering and Computer Science, The Catholic University of America, Washington, DC, United States

A general-purpose graphics processing unit (GPU) offers a powerful processing platform to accelerate non-graphics applications such as tensor estimation in Diffusion Tensor Imaging (DTI). Diffusion tensor maps are computed on a voxel-by-voxel basis by fitting the signal intensities of diffusion weighted images as a function of their corresponding b-matrices. This computation can be significantly accelerated by using the GPU. This study presents the application of using GPU hardware in diffusion tensor estimation by accelerating the weighted multivariate linear regression. The results are tested in simulated 3D brain dataset and show faster computation time against the CPU. The proposed GPU framework can accelerate DTI simulation and can be readily applied to quantitative assessment of the DTI using bootstrap analysis.

Klaus Hermann Fritzsche¹, Marco Nolden¹, Hans-Peter Meinzer¹, and Bram Stieltjes^1
1German Cancer Research Center, Heidelberg, Baden Württemberg, Germany

Q-ball imaging provides insights into aspects of brain structure in living humans that could not be studied previously. The lack of standardized and extensible software tools for I/O, reconstruction, interactive visualization, and statistics impedes development and sustainable evaluation. The diffusion imaging component of the Medical Imaging Interaction Framework (MITK-DI) aims at supporting cutting edge diffusion imaging techniques and, in contrast to most other frameworks, addresses all aspects of application design including full integration into an application platform and fluent workflows. MITK-DI therefore allows covering the complete cycle from raw-data to post processing and statistics.

Divya Kishore Singh Rathore¹, Sanjay K Verma², RKS Rathore², and Rakesh K Gupta^3
1Imaging R&D, ADISL, Kanpur, UP, India, ²Mathematics and Statistics, Indian Institute of Technology, Kanpur, UP, India, ³Departments of Radiodiagnosis, Sanjay Gandhi Post Graduate Institute of Medical Sciences, Lucknow, UP, India

A worldwide interest in research and development in MRI and associated protocols over last 2 decades has generated a need for tools and frameworks that allow rapid development of research grade software. We propose a plugin based architecture that allows developers to add extended support for various file-formats, post-processing algorithms, re-use of existing libraries etc. As an illustration, we are describing a plugin for post-processing of Diffusion Tensor MRI data.

Pierre Fillard¹, and Nicolas Toussaint^2
1Parietal Reseach Team, INRIA Saclay Île-de-France, Gif/Yvette, France, ²Imaging Sciences, King's College London, London, United Kingdom

Although the MedINRIA software has been out for three years, it has never been formally presented to ISMRM. The objective of this work is to expose the rationale for the development of MedINRIA, a clinically oriented medical image processing software, and to present its DTI functionalities. We present the processing pipeline going from DICOM to fiber tracts. Notably, it comprises a unique anisotropic smoothing procedure enhancing the tensor quality of datasets of moderate SNR typical of clinical data. We further describe the interactive tract-of-interest selection tool and show an example of fiber bundle extraction and analysis with MedINRIA.