摘要
肺区自动分割是肺部肿瘤计算机辅助诊断系统的关键之一。文章采用多阈值和区域生长方法,先去掉背景,再去掉气管/支气管,最后对提取出来的肺区使用滚球的方法进行修补。该方法速度快、人工干预少、准确。
Automated segmentation of lung is one of key techniques in computer aided diagnosis systems to detect lung tumor. The muhi-criterion segmentation algorithm and region growth algorithm to get rid of the background and the trachea and main bronchi are used at first. Then the lung fields are gained and repaired with falling-ball method. The method has high speed, little manual intervention and nicety.
出处
《科学技术与工程》
2008年第24期6459-6462,共4页
Science Technology and Engineering
关键词
CT图象
肺区分割
多阈值
区域生长
CT image lung segmentation multi- thresholding region growth
