A Vessel Segmentation Method for Low Contrast CT Angiography Image

CT angiography image vessel segmentation technique plays a critical role in many practical applications such as diagnosis of diseases and surgical planning. Blood vessels after angiography are often defective in low contrast and high noise due to individual differences and limitations of imaging device. A set of segmentation methods has been proposed for such data feature, which firstly adopts combination of average filtering with high frequency emphasizes filtering enhancement algorithms to perform enhancement preprocessing on the original image, thereby enhancing the contrast ratio of blood vessels and surrounding area; secondly, it performs vessel extraction on the enhanced data using region growth algorithm to obtain the segmentation data of richer vessel branches. Experimental results prove that the segmentation approach proposed here can accurately perform vessel segmentation and effectively avoid incorrect segmentation with better robustness.