模糊神经网络预测电针镇痛过程中生化指标变化的应用研究

目的针对电针镇痛过程中生化指标检验的时滞性、费用高等不足，本文提出基于模糊神经网络预测电针镇痛过程中生化指标变化的方法。方法模糊神经网络的性能通过调整隶属函数的类型和隶属函数的个数进行优化。模糊神经网络的训练数据来源于电针镇痛患者的ECG、EEG等生理电信号以及检测的生化指标。我们从中选取45个样本为训练样本，剩余的15个样本作为测试样本。结果从模型训练结果的变化来看，训练数据预测的结果比较接近实际检测的结果。结论模糊神经网络预测的结果与实际检测的结果非常吻合，这为电针镇痛过程中生化指标变化预测提供了一种新的方法和思路。

Biochemical index estimation from bioelectric signal information-the ANFIS method

Objective In this paper, a fuzzy neural network（FNN） is proposed for fusing the bioelectric information, and realizing the monitoring of the change of biochemical index. Methods ECG, EEG and tip perfusion index （TPI） data from 10 patients undergoing general electro-acupuncture analgesia were extracted and the fuzzy neural network was trained by 45 cases, and tested by the other 15 cases. The performance of ANSIF model is optimized as a function of a type of membership function and number of membership functions. Results Optimized ANF1S model well predicts the results compared with the experiment data. Conclusions Based on the established model, the analysis results of the process parameters show that the change of reference biochemical index can be estimated by monitoring bioelectric sianal parameters.