采用人工神经网络对肌肉痉挛进行定量评估

Quantitative assessment of muscle spasticity using artificial neural network

利用系统辨识方法研究脊髓多节段硬化病人与正常对照组的髌韧带反射特性 ,抽取一系列系统特征参数。在此基础上 ,探讨利用人工神经网络对测量的系统参数进行分级和评定的方法及效果 。

Patellar tendon reflexes were studied in normal subjects and multiple sclerosis (MS) patients. The input output properties of the system were characterized using several system parameters. Based on those parameters, an artificial neural network was used to assess the severity of spasticity. The results indicate the potential values of this method in clinical practice as an approach towards quantification of the tendon jerk, a key clinical measure of spasticity.