基于阈值和snake模型的三维医学图像自动分割  

3D Medical Image Segmentation Using Thresholding and Snake Model

作　　者：张谦[1] 耿国华[1] 周明全[1] 

机构地区：[1]西北大学可视化技术研究所,陕西西安710069

出　　处：《微机发展》2005年第2期109-111,114,共4页Microcomputer Development

基　　金：国家自然科学基金资助项目(60271032)

摘　　要：提出了一种利用snake模型和基于连通性阈值算法进行三维医学图像的自动分割方法。根据三维医学图像的特点,首先选取该图像的中间层图像,利用基于连通性的阈值算法对其分割;其次利用邻层图像分割结果和snake模型来指导下一层的图像分割。实验结果表明,该方法可以明显提高分割的准确率和速度。Propose a new 3D medical image automated segmentation method, which use snake model and connectivity-based threshold algorithm. The algorithm is based on specialty of the 3D medical image. First,make use of the connectivity-based threshold algorithm to segment the center slice. Second,utilize the result and snake model to guide segmentation of the next slice. Experiments show that the proposed method can obviously improve the accuracy of the segmentation and reduce the time needed.

关 键 词：三维医学图像 连通性阈值算法 smke模型 自动分割 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]