Automated sulcal segmentation using watersheds on the cortical surface.

摘要: The human cortical surface is a highly complex, folded structure. Sulci, the spaces between folds, define location on cortex and provide parcellation into anatomically distinct areas. A topic that has recently received increased attention segmentation of these sulci from magnetic resonance images, with most work focusing extracting either sulcal folds or curve representations sulci. Unlike methods, we propose technique extracts actual regions surround sulci, which call "sulcal regions." method based watershed algorithm applied to geodesic depth measure surface. well-known problem tendency toward oversegmentation, meaning single region segmented as several pieces. To address this problem, postprocessing merges appropriate segments algorithm. are then manually labeled by simply selecting mouse click preliminary study reported. Finally, scheme presented for computing complete