血管内超声图像序列的自动边缘检测

血管内超声成像可以显示血管内腔、管壁清楚的实时截面图像，从而使定量测量血管参数成为可能。传统手动追踪内腔和管壁边界的方法既费时又有较大的主观误差，本文基于快速主动轮廓算法模型提出一种适合血管内超声图像的自动边缘检测方法。实验结果发现，自动检测的内腔、血管壁面积与手动追踪非常吻合．具有较高的相关系数(0．94和0．95)和较小的系统误差(-0．32和0．56)，可作为医生可靠而准确的诊断工具。

Automated Contour Detection for Intravascular Ultrasound Image Sequences

Intravascular ultrasound can provide clear real-time cross-sectional images including lumen and plaque. Quantitative measurement can be performed by these images. Traditionally, the luminal border and medial-adventitial border are manually traced. This process is time-consuming and the subjective difference is larger. A new automated contour detection method is used for intravascular ultrasound images based on a fast active contour model. Experimental results show that luminal and vessel area measurements after automated detection are in agreement with manual trace with high correlation coefficients (0 94 and 0 95) and small system difference (-0 32 and 0 56). Thus it is used as a reliable and accurately diagnostic tool.