SUGAR: Spherical ultrafast graph attention framework for cortical surface registration

Jianxun Ren, Ning An, Youjia Zhang, Danyang Wang, Zhenyu Sun, Cong Lin, Weigang Cui, Weiwei Wang, Ying Zhou, Wei Zhang, Qingyu Hu, Ping Zhang, Dan Hu, Danhong Wang, Hesheng Liu

Abstract: Highlights•We propose a deep-learning framework for cortical surface registration.•A novel spherical graph neural network was developed for cortical spherical meshes.•Fold and distortion losses were used to preserve topology and control distortions.•Extensive evaluation was performed in over 10,000 scans from 7 diverse datasets.•Our method showed sub-second run time, superior accuracy, and minimal distortions.