基于人机交互的心理健康动态监测数据分析及系统设计

Dynamic monitoring of mental health based on human-computer interaction Data analysis and system design

虽然市场上有许多生理信号采集设备,用于人体心理健康监测。但这些设备不具有便捷、实时性等特点,不适合日常的推广。针对此问题,设计了一款基于心电、表面肌电的多模态生理信号采集系统,可用于对心理健康状况开展监测分析。文章包括以下内容:(1)系统硬件电路的设计,实现了便捷式、低功耗的多通道同步采集系统;(2)系统软件设计,包括设备驱动程序,设备应用程序和PC端应用程序;(3)系统数据预处理算法实现过程,并对心电和肌电信号进行预处理,并将采集到的生理信号波形与美国BIOPAC公司的MP150系统同等条件下采集的生理信号波形进行对照,保证多模态生理信号采集系统的有效性。最后,通过多分类SVM和DS证据理论算法对被试者的心理压力状态进行分类,实验证明多模态生理信号评估心理压力的可行性。

Although there are many physiological signal acquisition devices on the market,they are used for human mental health monitoring.However,these devices are not convenient,real-time and other characteristics,and are not suitable for daily promotion.To solve this problem,this paper designs a multimodal physiological signal acquisition system based on ECG and surface EMG,which is convenient for people to use in the dynamic detection data analysis of mental health.The content of this paper includes the following contents:(1)The design of the system hardware circuit realizes a convenient,low-power multi-channel synchronous acquisition system;(2)System software design,including device drivers,device applications and PC applications;(3)The system data preprocessing algorithm is implemented,and the ECG and EMG signals are preprocessed,and the collected physiological signal waveform is compared with the physiological signal waveform collected by the MP150 system of BIOPAC in the same conditions to ensure the effectiveness of the multimodal physiological signal acquisition system.Finally,we classify the psychological stress state of the subjects through multi classification SVM and DS evidence theory algorithm,and the experiment proves the feasibility of multimodal physiological signals to evaluate psychological stress.