摘要
在伽玛刀治疗计划系统中 ,为了给三维重建提供直观的参考依据 ,要求准确地对体表、病灶和其他组织进行轮廓提取。针对目前该系统中以手工轮廓提取为主的现状 ,本文提出并实现了一种新的图像轮廓提取算法。该方法利用初始轮廓逐步逼近最佳轮廓 ,不仅将图像灰度变化的微分信息作为边缘提取的依据 ,同时以图像轮廓的整体几何信息作为指导 ,从而在满足边缘检测精度要求的前提下 ,一定程度上解决了噪声对轮廓特征提取的影响。
In gamma knife treatment planning system, conto ur detection of the body surface, tumors and other tissues must be done precisely in order to provide intuitionistic reference grounds for three-dimensional recon struction. A new contour detection algorithm was introduced to replace the curre ntly predominantly used manual contour detection. This method was to close in up on the optimizing contour line on the basis of the initializing contour line. Be sides the differential information of the image gray level, the global feature o f the image contour was introduced to the process of edge detection so that edge detection and contour extraction could be integrated into one course and had hi gh precision of edge location.
出处
《机床与液压》
北大核心
2004年第6期39-40,46,共3页
Machine Tool & Hydraulics
关键词
伽玛刀
治疗计划系统
医学图像
初始轮廓
逼近
Gamma knife
Treatment planning system
Medical ima ges
Initializing contour
Close in upon
