利用EMG数据作自动诊断的人工神经网络专家系统

The Artificial Nerve Network Expert System for Auto-Diagnosis Based on EMG

目的 研究并实现一种基于EMG数据诊断腰腿痛的神经网络专家系统。方法 系统将专家系统的推理能力与神经网络的学习能力结合起来 ,利用专家系统中诸多临床医师检测经验与总体诊断结果和神经网络能对单块肌肉的肌电图进行计算分析来完成腰腿痛的诊断工作。结果 诊断腰腿痛的神经网络专家系统通过对其系统结构、系统原理及诊断方法进行阐述 ,体现该系统的严密科学性及临床实用性。结论 该系统设计科学 ,诊断快速、准确 。

Objective To develop an artificial nerve network expert system for diagnosis of lumbocrural pain based on electromyography(EMG) data. Methods The deduction ability of expert system and the study ability of network were integrated into this system, thus the diagnosis of lumbocrural pain would be extablished by the abundant examination experiences and general diagnostic results of a great many clinical practitioners in expert system and algorithm analysis for single muscle EMG with nerve network. Results The construction,principle and diagnostic procedures of the artificial nerve network expert system were described and the scientificalness and clinical practicability of this system were clarified. Conclusion The system is designed scientifically and it is accurate and efficient in diagnosis with EMG.