A simplified fuzzy connectedness method used for segmentation of vessel images

Vasculature is of utmost importance in planning surgical procedures and diagnoses of vessel diseases such as angioma and vessel stenosis. Fuzzy connectivity algorithm computes strength of connection between two voxels of the image and can be used to separate object from background of image. It needs to compute affinity of voxel of 4-connection in all paths, which is a big computing burden. In this paper, a simplified fuzzy connectedness method (SFC) is proposed. First, for the calculation of affinity, only the voxels satisfied certain condition are used. That reduces the computing range dramatically. Second, after careful analysis, only parts of the paths are searched which speeds up the segmentation greatly. Segmented results of both synthetic image and clinical image are very promising. The comparison between SFC and manual segmentation shows that the SFC algorithm improves efficiency and is potential for the segmentation of vessels in medical images.