基于随机游走算法半自动测量颈动脉狭窄度

Semi-Automatic Measurement for the Degree of Carotid Stenosis Based on the Random Walk Algorithm

目的应用随机游走算法半自动测量,提高CT血管造影(CTA)对颈动脉狭窄的诊断效率。方法回顾性分析2019年6月至2019年9月同期行头颈部CTA及脑数字减影血管造影(DSA)患者的影像学资料,应用随机游走算法半自动提取CTA原始横断面图像颈动脉轮廓及血管中高亮区域的轮廓,进一步应用阈值法得到二值化图像并获取两部分面积,最后计算出颈动脉狭窄度,并与DSA评估结果相比较。结果随机游走算法应用于CTA原始图像半自动测量颈动脉狭窄度,与DSA诊断结果具有高度一致性,一定程度上提高了传统CTA诊断轻、中度颈动脉狭窄的敏感性,对颈动脉重度狭窄的半自动诊断效能提升较大。结论随机游走算法是半自动诊断颈动脉狭窄的可靠方法,可提高传统CTA的诊断效率。

Objective To improve the diagnostic efficiency of computed tomography angiography(CTA) for carotid artery stenosis by semi-automatic measurement with random walk algorithm. Methods The imaging data of patients undergoing head and neck CTA and cerebral angiography DSA in the same period from June 2019 to September 2019 were analyzed retrospectively. The contours of carotid arteries and highlight areas in blood vessels were extracted from CTA original cross-sectional images and measured semi-automaticallyusing random walk algorithm. Furthermore, the threshold method was applied to obtain the binarized image and the areas of two parts. Finally, thedegreeofcarotid artery stenosis was calculated, which was compared with the results of DSA. Results Random walk algorithm was implemented in the semi-automatic measurement of carotid stenosis in CTA sources images, which was highly consistent with the diagnostic results of DSA.Itimproved the sensitivity of traditional CTA in diagnosing mild and moderate carotid stenosis to some extent and greatly improved the efficiency of semi-automatic diagnosis of severe carotid stenosis. Conclusion Random walk algorithm is a reliable method for semi-automatic diagnosis of carotid artery stenosis, which can improve the diagnostic efficiency of traditional CTA.