基于神经网络的案例推理医疗诊断

在医疗诊断中，案例推理是一种有效的思维方式，但在实际应用中，存在许多难点，如大型案例 案例的表示和相似案例高效检索问题，结合前向神经网络的特点，来实现案例检索，为减少多层前向网络进行局部极小的机会，改进网络的性能，提出了神经元和前向神经网络集成，构成一种增强型神经网络结构，数学和仿真证明这种网格能减少进入局部极小的机会。此外，借助于数据库技术和多级索引方法，设计了一个医疗诊断专家系统，结果表明上述

Case-Based Reasoning Medical Diagnosis Based on the Strength Neural Network

Case Based Reasoning (CBR)is an effective method in medical diagnosis. But there exist some problems such as presenting of cases and fast searching for similar cases in large case bases, when it is applied. In this paper, foreword multi layer networks are used for CBR because of their advantages. In order to decrease the chance of MLN's entering local minima and improve its effectiveness, a strength neural network is also proposed with a backward propagation neural network and an adaptive neuron integrated. And mathematical proof and simulation result are also given. Moreover, by means of the database technology and multi level index,the design of a medical diagnosis system proves that choosing the network to retrieve cases in medical diagnosis is valid.