中文摘要：

为提高医疗辅助诊断系统的检索效率，提高诊断准确率，降低误诊率，建立了基于贝叶斯学习与基于案例推理方法的信息融合模型。通过CBR方法进行案例检索，以贝叶斯网络作为数据挖掘工具，由病例样本获得诊断属性间的关系，建立CBR相似评价函数。CBR通过相似评价函数在病例库中检索与目标病例最相似的结果。通过对心脏病的医疗诊断实例对模型应用实现效果加以分析及验证。

英文摘要：

To raise the rate of accuracy and reduce the misdiagnosis rate in computer aided medical diagnostic system, a hybrid model based on Bayesian networks learning and case based reasoning was presented. In this model, CBR（cased based reasoning）was adopted as the method to index cases, and Bayesian network was used as the tool of data mining. Relations of diagnosis attributes were obtained from case samples,and similarity evaluation function of CBR was built. With the similarity evaluation function, CBR index the case database, so that the most similar case could be obtained. Then, with a case of heart disease diagnosis, the effect of the hybrid model was illuminated.