基于反向传播神经网络和医疗物联网的心理压力评估方法

Psychological pressure detection system based on BP network and IoMT

针对基于脉搏信号的人体心理压力测量分类结果精度不高的问题,提出一种基于反向传播(BP)神经网络和医疗物联网(IoMT)的心理压力检测系统设计方法。该系统采用以MAX30101为核心的光电容积脉搏波(PPG)采集模块来获取人体PPG数据,并基于小波去噪和RR尖峰检测方法得到心室搏动间距,利用时域和频域上的处理提取出心率变异性(HRV)作为特征,设计Adam自适应优化算法训练的神经网络模型实现心理压力程度3分类。同时为实现长时间不间断的检测与分析,设计基于IoMT的心理压力检测数据远程平台,实现24h连续用户健康检测。实验结果表明,该系统检测准确度达到88.7%,较传统的心理量表测评和激素测量法能够有效方便的对心理压力程度进行分类。

In order to improve the accuracy of classification results of human psychological pressure measurement based on pulse signal,a psychological pressure detection system based on back propagation network and internet of medical things(IoMT)is proposed.The system uses the photoplethysmography(PPG)acquisition module with MAX30101 as the core to acquire human PPG data,and obtains ventricular beat-to-beat interval based on wavelet denoising and RR peak detection.The heart rate variability extracted by processing in time and frequency domains is taken as the feature,and a back propagation neural network model trained by Adam adaptive optimization algorithm is established to classify the degree of psychological pressure into 3 categories.At the same time,a remote platform based on IoMT for psychological pressure detection is designed for 24-hour continuous user health monitoring,thereby achieving long-term uninterrupted detection and analysis.The experimental results show that the system has a detection accuracy of 88.7%,and that it is more effective and convenient to classify the degree of psychological stress than traditional psychological scale and hormone measurement.