A Physical Basis for Multi-Fiber Reconstruction from DW-MRI Data

Recently various mathematical models have been proposed to model the signal attenuation obtained from Diffusion Weighted Magnetic Resonance Imaging (DW-MRI). Though effective to various extents, almost all of the existing methods involve model parameters which are abstract mathematical quantities without any tangible connection to physical quantities (e.g. the b-value, gradient pulse duration, pulse separation etc.) involved in the DW-MRI acquisition process. To address this disconnect, in this paper, we present a multi-compartmental model which uses a physical model for restricted diffusion in the cylindrical geometry as the constituent basis function for multi-fiber reconstruction. Through extensive experiments on synthetic data we establish the superiority of the proposed method over the state-of-the-art techniques in terms of fiber orientation detection accuracy. We also present detailed results using human and rat brain data and demonstrate that our method leads to meaningful m...

• Type: Conference
• Year: 2009
• Where: ISBI
• Authors: Ritwik Kumar, Angelos Barmpoutis, Baba C. Vemuri, Paul R. Carney, Thomas H. Mareci
• Keywords: ISBI 2009, Medical Imaging, Methods Involve Model, Model, Multi-fiber Reconstruction