Study And Application Of Anti-motion Artifact Method For Wrist-type PPG Signal

Optical sensors are used to obtain cardiovascular physiological signals from the outside of human skin through photoplethysmography(PPG).Through analysis and calculation,physiological parameters such as heart rate can be effectively and non-invasively monitored and used for disease diagnosis.Cardiovascular diseases are often hidden,so wearable devices with integrated PPG sensors bring new hope to more convenient and fast health monitoring.The detection device based on the wrist-type PPG signal coincides with people’s cognition of wearable devices,and can promote and improve people’s cognition of their own health.This has brought huge economic benefits to the related industries of wearable devices,and also provided important reference value for the popularization of medical devices in the future.However,placing the PPG sensor on the wrist to detect large artifact interference makes the obtained heart rate and other physiological parameters inaccurate.At present,most PPG sensors with a wavelength of about 500 nm are used to obtain higher-quality pulse wave signals,but they are not suitable for detecting parameters such as blood oxygen saturation,and motion artifacts still exist.Therefore,based on the wearable device,this paper designs a wrist PPG signal acquisition system based on a reflective near-infrared sensor.The system uses STM32L4R9VIT6 with ultra-low power consumption;the PPG sensor MAX86150 uses near-infrared dual-wavelength LEDs with LED wavelengths of 660 nm and 880 nm,which can be used to detect blood oxygen saturation;MPU9250 is used to obtain acceleration signals as motion Reference sources of noise for artifacts;the display part uses an ink screen,which can greatly reduce power consumption.After the system design was completed,the wrist-type PPG signal was synchronously collected and analyzed.It was verified that the wrist-type PPG signal acquisition system using near-infrared light can collect pulse wave data and calculate the blood oxygen saturation.The use of least mean square(LMS)adaptive filtering algorithm based on motion noise reference to remove noise artifacts can have a certain anti-artifact ability for mild motion.The wrist-type PPG signal acquisition system was compared with Mindray PM-9000 medical monitor.Eight volunteers were tested on the static oxygen saturation,heart rate in the squat state,and blood oxygen saturation in the squat state,and a total of 72 sets of experimental data were obtained.The maximum absolute error of the blood oxygen saturation in the static test in the first two items is 2%.The maximum absolute error of heart rate in the squat state is 3 beat per minute,and the average value is 1.2 beat per minute.The final absolute error of the blood oxygen saturation test result is 4%.The consistency analysis based on the Bland-Altman method was done for the above three results.The results show that the sample data of the first two items are within the 95%consistency limit confidence interval,and 8% of the sample data of the last item falls outside the 95% consistency limit confidence interval.Explain that the wrist-type PPG signal acquisition system designed by the text has a certain anti-artifact effect and reliability under light movement.However,due to the presence of artifacts,the measurement of blood oxygen saturation under exercise is still difficult to obtain subtle changes in oxyhemoglobin and deoxyhemoglobin,and further research and improvement are needed.