摘要： 针对具有强背景噪声及伪影干扰的血管内超声图像，提出一种自动提取血管壁内外膜边缘的新方法。利用图像序列的空间和时间相关性对噪声进行抑制，在GVF-Snake模型中引入调节因子和自适应法向外力，扩大边缘的捕捉范围，以提高活动轮廓对噪声的鲁棒性和提高图像提取精度，同时利用三次B样条增强边缘平滑性，加快收敛速度。实验结果表明，该方法的检测准确性较高，运行时间较短。

关键词: 血管内超声图像, 边缘检测, 活动轮廓模型, 梯度向量流, 自适应外力, 三次B样条

Abstract: Aiming at the Intravascular Ultrasound(IVUS) image with strong background noise and pseudo shadow interference, a novel method of automatic vessel wall edge detection for IVUS images is proposed. In the process of detecting edge, the noise can be reduced by using the spatial and temporal correlation among the sequences. The new regulatory factor and adaptive normal external force are introduced in the GVF-Snake model which not only enlarges the edge capture range and makes the extraction effect of the image more accuracy, but also improves the robustness of the active contour to noise. It uses cubic B-spline to enhance the edge smoothness, and speeds up the convergence. Experimental results demonstrate that the method has higher accuracy and shorter running time.

Key words: Intravascular Ultrasound(IVUS) image, edge detection, active contour model, Gradient Vector Flow(GVF), adaptive external force, cubic B-spline

中图分类号: 

• TP391