基于三维CTA血管拉直的腹主动脉瘤自动定位方法

Automatic AAA Positioning Method Based on Three-dimensional CTA Vascular Straightened

目的:腹主动脉瘤越来越严重地威胁着人类健康,CTA成像设备已经成为该疾病的重要诊断工具,从海量的三维CTA影像数据中自动定位腹主动脉瘤位置对疾病辅助诊断系统有重要意义。方法:首先,对分割出的腹主动脉及其分支血管进行三维拓扑细化,提取血管骨架线;其次,利用A-Star最优路径算法确定主动脉中心线,并进行平滑处理;再次,沿主动脉中心线重建血管切面,拉直主动脉;最后,计算主动脉横截面积,定位瘤体位置。结果:经过4例由GE 16层螺旋CT采集的腹主动脉瘤CTA数据的验证,本方法能够100%检测到瘤体,自动定位的瘤体所在原图像中的层数与人工标注的重合率达到90%以上。结论:本文方法可以从海量的腹主动脉CTA数据中自动定位瘤体位置,具有较强的临床应用价值。

Objective： Abdominal aortic aneurysm （AAA） is a serious threat to human health, and the CTA imaging has become an indispensable diagnosis means. It＇s very important to automatically locate the position of the AAA from the flood of 3D CTA images for such aided diagnosis systems. Methods： First, 3D topology thinning algorithm is used to extract skeleton of the abdominal aorta and its branching vessels. Secondly, A-Star optimal path algorithm is used to find the aortic centerline, then smoothing it. Thirdly, do reconstruction of vascular cross-sectional slice along aortic centerline to straighten aorta. Finally, calculate the aortic cross-sectional area to locate the tumor. Results： Four cases of AAA CTA data gathered by GE 16-slice spiral CT acquisition tested, the present method can detect all the tumor, and the coincidence rate of automatic positioning and artificial tagging the layers of tumors in the original image is above 90%. Conclusions： This method can automatically locate the tumor position from massive abdominal aorta CTA data, and has a strong clinical value.