为识别CT图像中的肺结节,提出了一种结合规则和支持向量机（SVM）的识别方法,来对分割出来的感兴趣区域（ROI）进行分类.该方法首先计算候选ROI的形态特征,利用基于规则的方法筛去非结节的区域;然后把筛选之后剩余的候选ROI作为测试样本和训练样本,计算它们的灰度和纹理等特征;最后把灰度、形态和纹理特征值作为SVM的输入,对经过基于规则筛选之后剩余的ROI进行分类.实验结果表明：基于规则的方法虽然没有漏检,但误判的可能性最大;结合规则和SVM的方法漏检的可能性要比SVM方法漏检的大,但误判的可能性小.

In order to effectively recognize lung nodules in CT images,a recognition method combining the rule-based method and the support vector machine（SVM） is proposed to classify the regions of interest（ROIs）.In this method,first,shape features of candidate ROIs are calculated,and some non-nodule regions are filtered out by using the rule-based method.Then,the remaining candidate ROIs are taken as testing and training samples,whose grayscale and texture features are calculated and used as the inputs of SVM to classify the remaining candidate ROIs.Experimental results show that the rule-based method may result in high possibility of misdiagnosis although there is no nodule omission,while the proposed method combining the rule-based method and the SVM is of low possibility of misdiagnosis but of obvious nodule omission.