Automated shape analysis of cerebral vessels from MR angiography images as a biomarker of atherosclerosis

点击翻译按钮获取中文摘要

The aim of this study is to develop a biomarker of atherosclerosis by computer-assisted shape analysis of cerebral arteries from MRA data.First, we analyzed pilot data of 10 cases using multiscale morphological approach. We defined a feature value that caused a significant difference between a control group and an atherosclerosis group, as “shape regularity index of vessel (SRIV)”.Second, 51 cases were selected as a validation data set and divided into three groups by visual evaluation of vessel wall (Grade 0: no irregularity, Grade 1: mild irregularity, Grade 2: severe irregularity). A significant difference was found in SRIV value between Grade 0 and 1, and between Grade 1 and 2.We concluded that SRIV value had a correlation with vessel wall irregularity evaluated visually, and this value would be an objective biomarker of atherosclerosis.