Abstract #0718

Xu D, Hao X, Plessen K, Bansal R, Peterson B, Geng W

Diffusion tensor imaging provides information on the connectivity of white matter fibers in the brain, which is undetectable using other imaging modalities. In most applications, DTI data are analyzed independently from other imaging modalities, with regions of interest (ROIs) for fiber tracking identified manually by experts. However, the accuracy of manual segmentation can be unreliable because of morphometric and functional variations between individual brains. Moreover, the procedure is time consuming and subjective. Based on our previous work on image co-registration across modalities and neural fiber tracking, we propose an automated framework that uses data from multiple imaging modalities to identify ROIs for fiber tracking. We demonstrate the effectiveness of our procedure through application to a population of Tourette’s syndrome (TS) patients and a group of age-matched normal controls. Our procedure requires minimal human intervention and is therefore relatively objective and accurate.