Vascular centerline extraction of CTA images based on minimal path and Bayesian tracking

Vascular centerline and centerline branch information are especially important for the analysis of vascular disease and vascular intervention surgery. Minimal path propagation with backtracking algorithm (MPP-BT) has a good performance of vascular centerline extraction, and solves endpoint problem, shortcut problem and accumulation problem of traditional minimum path algorithm. Given a path through two endpoints, the MPP-BT is used in this paper to extract vascular centerline by manually setting a starting point; however, the resultant centerline lacks branch information. To obtain centerline branch information, a new method is proposed, which is based on minimum path and Bayesian tracking. The process includes three steps: Firstly, a new multiscale filtering algorithm is used to enhance vascular structure; secondly, vascular centerlines are extracted using minimum path; finally, quantized information of centerline and branch structure is obtained by Bayesian tracking. Experimental results show that the proposed method can effectively extract vascular centerline with the corresponding branch information.