The Near-Infrared Forearm Vessel Image Segmentation and Application Using Level Set

A robust level set-based near-infrared forearm vessel image segmentation method is proposed. First, a near-infrared forearm image capture device is developed. More than 500 forearm images are accumulated by using this device. Second, an image segmentation algorithm is designed. The multiscale second order local structure of image is considered to enhance the vessel edges. The distance regularized level set evolution approach is used to implement the vessel segmentation. Third, the proposed algorithm is employed to assist the searching of the potential Arteriovenous Fistula Thrombosis (AFT). Many experiment results have shown the proposed system and method can locate the superficial vessels of forearm correctly; and the recognition rate of AFT is better than 95%.