Xiujuan Geng1, Hong Gu1, Wang
Zhan2, Wanyong Shin1, Yi-Ping Chao3, Norbert
Schuff, Ching-Po Lin4, Yihong Yang1
1Neuroimaging, National Institute on
Drug Abuse, NIH, Baltimore, MD, United States; 2Department of
Radiology, University of California, San Francisco; 3Department of
Electrical Engineering, National Taiwan University, Taiwan; 4Institute
of Brain Science, National Yang-Ming University, Taiwan
We
propose a novel diffeomorphic registration method for diffusion MRI data by
mapping their orientation distribution functions (ODFs) represented with
spherical harmonics (SHs). The registration is based on optimizing a
diffeomorphic demons cost function. ODF reorientation is performed by
rotating the SH coefficients to maintain the consistency with the local fiber
orientations. Rotation on SH coefficients avoids the estimation of principle
directions which has no analytical solution and is time consuming. The
performance was tested using different SH orders. Results show that
registration with higher orders improves the registration accuracy in terms
of smaller similarity error and higher directional consistency.