研究目的
To develop an integrated non-invasive vital signs monitoring system to measure vital signs such as heart rate variability and alterations in skin temperature, and to assess the ability of the system to detect stress.
研究成果
The developed integrated non-invasive vital signs monitoring system was validated to accurately measure pulse rate and detect stress based on changes in pulse rate and skin temperature. The system represents a feasible method for stress detection in daily life, though further improvements are needed to handle motion artifacts and individual differences in stress responses.
研究不足
The system's accuracy may be affected by motion artifacts, requiring subjects to remain still during measurements. The study did not account for rapid acclimatization to stress stimuli in some participants, which could affect stress detection accuracy.
1:Experimental Design and Method Selection:
The study utilized near infrared (NIR) and far infrared (IR) cameras to measure pulse rate and skin temperature changes non-invasively. An algorithm was developed for high-accuracy pulse rate measurement from NIR images of the human cheek, and skin temperature was measured from far IR images of the nose and forehead.
2:Sample Selection and Data Sources:
Five participants (23–26 years) for pulse rate measurement and four participants (23–25 years) for stress detection were involved. Informed consent was obtained prior to the trials.
3:List of Experimental Equipment and Materials:
NIR camera (Xbox One Kinect Sensor, Microsoft), far IR camera (OWLIFT Type-A, Infinitegra), and an ECG measurement circuit (EVAL – AD8232, Analog Devices) for validation.
4:Experimental Procedures and Operational Workflow:
Participants were seated facing the cameras, with pulse rate and skin temperature measured in relaxed and stressed states (induced by cold water applied to the wrist).
5:Data Analysis Methods:
Pulse rate was calculated from the interbeat interval (IBI) of the BVP waveform, and skin temperature changes were analyzed as the difference between nose and forehead temperatures. Stress detection was performed using SVM classification with 5-fold cross-validation.
独家科研数据包,助您复现前沿成果,加速创新突破
获取完整内容