Segmentation of Cerebral Venous Vessel in SWI Based on Multi-adaptive Threshold with Vessel Enhancement and Background Effects Elimination

Accurate segmentation and extraction of the cerebral veins in susceptibility weighted MR image is helpful to clinical diagnosis. This paper demonstrates a multi-step combination approach, which uses region dividing method to suppress the high-frequency background effects caused by in homogeneity in the bottom brain, utilizes Vessel Enhance Diffusion filter to enhance the continuity of the vein and suppress the nucleus area, and develops a multi-adaptive threshold method to combine tubular structure based method and gray scale based method to finally realize the accurate segmentation of the vein. The results showed that the approach can detect the very small vessels of the brain and segment the vein from the brain tissue effectively.