DEFORMABLE REGISTRATION OF DTI AND SPGR IMAGES

Registration of DTI data with structure data, such as SPGR data, has import application in quantitative analysis of brain microstructures such as tissue diffusivity. However, due to potential problems such as EPI geometric distortion, partial volume effect and image reslicing errors, accurate registration of these two types of MRI images is challenging. In this paper, we present a novel deformable registration approach for DTI data and SPGR data. The registration approach includes two steps: (1) brain tissue segmentation based on seven-channel fusion of DTI data in order to obtain a complete tissue segmentation map of CSF, GM, and WM in the DTI space; and (2) hybrid volumetric and surface registration of DTI and SPGR data with the tissue segmentation maps in both DTI space and SPGR space. Experimental results are provided to demonstrate the performance of our approach