基于可变模型的腹主动脉瘤分割算法

A Deformable Model-based Segmentation of Abdominal Aortic Aneurysms

提出了一种基于可变模型的腹主动脉瘤分割算法。它利用相邻两个切片间的相关性以及医学先验知识,可以快速地分割腹主动脉瘤CTA图像序列,并且结果与专家人工分割结果接近。本文将像素点分为三类:边界内,边界点,边界外,这样就直接决定了模型形变方向,大大提高了分割效率。

A novel approach for automatic segmentation of abdominal aortic aneurysms (AAA) based on deformable model was proposed in this paper. Combined with medical knowledge and relativity between two CT slices, the AAA could be segmented rapidly from a series of CT angiography (CTA). The results are close to the manual results by specialists. The proposed method uses a three-type class that distinguishes boundary, inside and outside non-boundary points, and in this way steers the deformation direction directly. The segmentation efficiency is greatly increased.