A segmentation algorithm of cerebral cortex based on dual edge detections and regional growing
Wen Li-li · Zhongguo jieru yingxiang yu zhiliaoxue · 2012
Objective To design a automatic segmentation algorithm for digital cross-section slice images of human brain,in order to segment the cerebral cortex accurately.Methods Cerebral cortex was obtained and extracted from gray matter using RGB space edge detection,Canny operator detection and regional growing algorithm.A comparative analysis between the segmentation results of automatic algorithm and manual method was performed.Results Highly coincident to that segmented by manual method,cerebral cortex segmented by automatic algorithm was characterized by complete,clear and smooth edge.Conclusion This algorithm is suitable for the segmentation of cerebral cortex in digital cross-section slice images.