基于启发式A^＊算法的超声图像颈动脉内膜提取

从超声图像准确提取颈动脉内膜,为基于颈动脉超声图像判断动脉粥样硬化服务。方法提出一种基于启发式A*算法从超声图像中提取颈动脉内膜边缘的方法。先使用图像分割法区分血管腔和血管壁,再采用结合图像灰度值特点的A*算法准确地提取颈动脉内膜边缘。结果通过对临床采集的32幅颈动脉超声图像的分析研究,表明本方法自动提取的结果与医生手工描绘的结果基本吻合。结论本方法有望应用于超声图像颈动脉内膜的自动提取。

Carotid Intima Extraction from Ultrasonic Images Based on Heuristic A* Algorithm

Objective To accurately extract carotid intima from ultrasonic images is the prerequisite of the arteriosclerosis diagnosis from ultrasonic carotid images.Methods A method based on heuristic graph searching algorithm (A algorithm) is proposed to detect carotid intima from ultrasonic images. First, the lumen and wall are divided by an image segmentation method. Then, A algorithm incorporating characteristics of the image intensity is applied to extract carotid intima. Results The experiments on 32 carotid ultrasonic images have shown that results of this proposed method match with those obtained manually. Conclusion This method can be efficiently used to automatically extract carotid intima from ultrasonic images.