中文摘要：

根据脑转移瘤在早期诊断过程中存在的问题,提出基于3D模板匹配的脑转移瘤计算机辅助检测方法。首先,建立合适的脑转移瘤模板;然后,在MR序列图像数据中寻找与模板相似的区域,相似度较高的区域很可能就是脑转移瘤。选取脑转移瘤为8 mm以下的12个病例（共有27个脑转移瘤）对本方法进行测试,其总体检测正确率达到81.5%,假阳性率为每层图像0.037 8个。实验结果表明,所提出的方法能为脑转移瘤的早期诊断奠定良好的基础。

英文摘要：

Early diagnosis for brain metastases has been one challenge issue in cancer therapies.This paper presented a computer-aided detection method for brain metastases based on the 3D template matching.First step,build up an appropriate brain metastases template;and secondly,search for regions of interesting（ROI） that similar to the template,which are very likely to be the candidate brain metastases.12 datasets that have total 27 brain metastases（less than 8 mm） were used to examine the efficiency of the method.The experimental results showed that the sensitivity of the method was 81.5% and the false positive was about 0.453 8 for every slice.Therefore,the computer-aided detection method provides a solution for the early diagnosis of brain metastases.