A Novel Method for Vessel Segmentation and Automatic Diagnosis of Vascular Stenosis

A novel method of vessel segmentation in angiocardiography image processing is studied. The process of vessel segmentation is implemented based on Hessian matrix, in which a new vascular function is employed. Subsequently, the vessel contour is extracted from the handled image by using the dual-stage region growing. Extracting vessel skeleton, searching skeleton point and measuring diameter are carried out to realize the automatic diagnosis of vascular stenosis based on vessel segmentation. The experimental results show that the stated method for vessel segmentation manages to accurately extract the main branch contour, and even most small vessels. The automatic diagnosis of vascular stenosis can get relatively accurate results, which may provide auxiliary references and quantify the judgment of vascular stenosis for doctors.