基于区域特征的中心点追踪的颈动脉分割

Carotid Artery Segmentation Based on Center Point Tracking of Regional Features

颈动脉狭窄可能会引起缺血性脑卒中,早期识别颈动脉血管形态对于评估脑卒中发生风险具有重要意义。通过对颈动脉三维重建的研究,可以还原颈动脉的结构,对于研究脑血管疾病的成因及临床诊断与治疗有很大的帮助。本文针对CTA图像的特点,在管状结构CT值筛选和多尺度滤波图像增强的基础上,采用一种基于区域特征的中心点追踪的分割方法,并在2套CTA图像上进行了实验。初步结果表明,算法对于颈动脉血管大部分二维图像能够较好地分割。

Carotid stenosis may cause ischemic stroke. Early identification of carotid vascular morphology is important for assessing the risk of stroke. Through the study of three-dimensional reconstruction of the carotid artery, the structure of the carotid artery can be restored, which is of great help to study the causes and clinical diagnosis and treatment of cerebrovascular diseases. In this paper, based on the characteristics of CTA image、the CT value screening of tubular structures and multi-scale filtered image enhancement, a segmentation method based on regional feature tracking is used, and experiments are carried out on two sets of CTA images. Preliminary results show that the algorithm can segment the majority of the two-dimensional images of the carotid artery.