Author : Angela Muhanga
Publisher :
ISBN 13 :
Total Pages : 118 pages
Book Rating : 4.:/5 (92 download)
Book Synopsis Network Modeling Based on Diffusion Tensor Imaging to Investigate Structural Integrity of Brain White Matter by : Angela Muhanga
Download or read book Network Modeling Based on Diffusion Tensor Imaging to Investigate Structural Integrity of Brain White Matter written by Angela Muhanga and published by . This book was released on 2013 with total page 118 pages. Available in PDF, EPUB and Kindle. Book excerpt: Traumatic Brain injuries (TBI's) affect approximately 1.7 million people in the US annually accounting for almost a third of injury related deaths. Diffusion Tensor Imaging (DTI) is a noninvasive method used to image fibrous tissue in the body. It is popular for use in clinical brain imaging in 3D because it can more effectively diagnose and track the recovery of TBI than other tools available. We focus on the analysis of structural integrity of fibrous tissue in the nervous system, particularly, brain white matter. In this paper, we outline two approaches for constructing network models from streamlines generated from brain DTI data; Graph-theoretic approaches and Network flow modeling. These approaches have been found useful in the field of computational neuroscience. They were chosen based on previous research methods described in Zhang, Correia and Laidlaw (2008). The objective of the thesis is to apply network models to DTI data. In this paper we construct a network model from streamline data. Future research is aimed at 1) analyzing the network using graph-theoretic approaches and 2) analyzing the network using network flows modeling. Other future goals include the exploration of other potential data computation and analysis tools that allow clustering, for example MUSKETEER whose developers published their research in 2012. Network approach to DTI data analysis is an attempt to summarize the data in a meaningful way. There's potential for developing algorithms that allow for automation and insights through comparison of different brain data both healthy and those in different pathology states. We applied an algorithm and tested it in MATLAB. The result was a reduction of streamlines from 5225 to 1052. The conclusion is that this algorithm is a feasible step in the right direction. The research could be made more useful by making data comparisons between multiple samples of brain data, from both healthy and TBI affected brain scans to provide more insight on how network models can applied to assess structural integrity of neural networks.