Abstract

In this paper, a single channel independent component analysis(sICA) is incorporated with Q-ball imaging (QBI) to enhance theperformance of diffusion orientation distribution function (dODF) reconstructionin diffusion weighted MRI (DW-MRI). In the proposed approach,the sICA is employed to decompose a dODF initially estimatedby analytical QBI. The initial estimated dODF is considered as a mixedsignal containing noise, artifacts and the desired “true” ODF to be extractedby sICA. The results from the numerical simulation demonstratethat the proposed sICA framework not only successfully separate thedesired ODF from the noisy data, but also achieves better angular resolutionthan analytical QBI and spherical deconvolution (SD).