声发射信号诊断的膝盖骨性关节炎

Acoustic Emission Signals in the Diagnosis of Knee Osteoarthritis

针对目前膝盖骨性关节炎(OA)发病率高,传统的诊断方法缺乏便携性,价格贵,存在放射性危害等局限性。该文提出了一种基于遗传算法优化BP神经网络的膝盖OA诊断。该方法以健康组和患病组为研究对象,将采集到的声发射(AE)信号通过遗传优化算法得到BP网络的最佳初始权值和阈值,利用BP神经网络对AE信号进行诊断分类。经实验室建立的膝盖OA声发射数据库的实验结果表明,该诊断方法效果好,误差小,能对膝盖OA患者进行实时检测和早期预测,具有临床价值。

Since the incidence of knee osteoarthritis （OA） is high, and the conventional diagnostic methods are non-portable, expensive and radiation damage etc. This paper proposes a diagnosis method of knee OA, which based on genetic algorithm to optimize the BP neural network. The proposed method takes the healthy group and disease group as the research object, and then makes use of the collected acoustic emission （AE） signal through genetic al- gorithm optimal BP network to obtain initial weights and threshold values, the BP neural network is used to classify the AE signals. The results of experimental knee OA acoustic emission database established by the laboratory show that the diagnostic method has well effect, small error, clinical value and to be able to real-time detection and early prediction of knee OA patients.