Abstract: To research on the application of computer-aided diagnosis expert system based on machine adaptive learning.To develop model on syndrome types for acute cerebral infarction and design the system in Visual Basic program language，then to apply the model in the clinic diagnosis.it has got the satisfied degree of similarity in the test of the 145 cases the authors collected from neurology department of the First Attached Hospital of Shanxi Medical University.In the following practical use，it produces appropriate diagnosis about 90.7%.It could effectively simulate the specialists diagnosis.

Key words: fuzzy math, computer-aided diagnosis expert system, object parameter

摘要： 对基于自学习功能的计算机辅助医学诊断专家系统的应用进行研究。应用基于模糊数学的方法建立急性脑梗塞早期分型医学诊断模型，利用VB语言进行系统的开发，并应用于临床诊断。对山西医科大学神经科已确诊的典型病倒145份进行诊断，计算机与实际诊断结果具有较好的贴近度。在随后对86例门诊患者的预测诊断中，诊断正确率达到了90.7%。该诊断模型与实际诊断结果具有良好的拟好的效果。

关键词: 模糊数学, 计算机辅助诊断, 目标参数