Segmentation of the Aortic Dissection from CT Images Based on Spatial Continuity Prior Model

Abstract

In order to improve the segmentation and reconstruction effect of the aortic dissection diagnostic equipment in hospital, we plan to develop a better three-dimensional reconstruction system of the aortic dissection to meet the requirements of the clinicians. This paper mainly introduces a series of preliminary work for the system: we utilize GVF snake model for descending aorta segmentation, extracting the aortic dissection membrane with the help of the Hessian matrix and the spatial continuity prior model based on Bayesian theory. We carried on the experiment in a series of continuous CT images, and the segmentation results were compared respectively with the manual segmentation results and the results without using the spatial continuity prior model. The experiment has proved that the spatial continuity prior model is effective for accurate segmentation of aortic dissection.