基于改进C-V模型超声颈动脉图像内-中膜厚度的测量

Measurement of Intima-Media Thickness in Ultrasonic Carotid Artery Images Based on Improved C-V Model

颈总动脉内-中膜厚度是心脑血管疾病早期筛查和诊断的主要标准之一。文章提出了一种基于超声图像的颈动脉内-中膜厚度半自动测量方法。最初在相关的颈动脉超声图像中选择感兴趣的区域,然后采用改进的C-V主动轮廓模型算法,在公式中加入局部强度信息,从而实现了图像局部结构的分割。该算法分为两步,第一步利用改进的C-V方法从超声图像中分割腔-内膜边界。在得到腔-内膜边界后,第二步使用类似的方法检测中-外膜边界。成功分割出腔-内膜和中-外膜边界后,可方便地计算内-中膜厚度。实验结果表明该方法可以在较少的人工输入并且不需要任何预处理步骤的条件下准确地测量内-中膜厚度。

The intima-media thickness of common carotid artery is one of major criteria for early screening and diagnosis of stroke.In this paper,a semi-automatic measurement method of carotid intima-media thickness based on ultrasonic images is proposed.The initial step is to select the interesting area in the relevant carotid ultrasound images,adopt an improved C-V method,and add local intensity information into the formula of variational level set,thus realizing the segmentation of local image.The algorithm is divided into two steps.In the first step,the improved Chan-Vese method is used to segment the inner boundary of lumen and intima from the ultrasonic image.Then we used a similar method to detect the boundary between the media and adventitia.The intima-media thickness can be calculated conveniently after lumen-intima and media-adventitia is detected.The results show that our method can accurately measure inner-medium thickness with less manual input and without any pretreatment steps.