中文摘要：

中心线的提取速度是提高结肠计算机辅助检测的效率的重要因素．为此提出了一种基于生成树的中心线快速提取算法（FMST）．在分析了最大生成树中心线提取算法（MST）存在大量冗余数据特点的基础上，利用边主源辅的搜索策略，保留趋于物体中心的关键点，通过消减冗余数据的方法来提高MST算法的速度．在10套已知中心线金标准的结肠仿真数据和2套结肠CT数据上的实验结果表明，FMST算法加快了MST算法的提取速度并且保持了中心线提取的准确性；在仿真数据上，FMST算法较MST算法的速度提高了80％以上，中心线重合率达到96．98％．

英文摘要：

Centerline Extraction is one of the crucial components in order to improve the performance of computer aided detection of colon lumen. In this paper, a fast centerline extraction algorithm based on Maximal Spanning Tree （FMST） is proposed. Specifically, FMST cuts off the boundary voxels to speed up the traditional Maximal Spanning Tree （MST） algorithm. Furthermore, based on searching strategy of boundary points primary and sourcepoint secondary, the key points which trend to the centers are kept, while the performance of the algorithm is improved by redundancy points reduction. The experimental results on 10 simulated colon models with centerline standard and 2 colon CT image series show that FMST is much faster than MST （speed up 80%） while preserves the accuracy of centerline （with 96.98%overlap rate）.