Study On Key Techniques Of Non-contact Detection Of Cardiovascular Physiological Parameters

Cardiovascular disease is a serious threat to human health.The key issue is the early detection and treatment.As we know,early detection of cardiovascular disease can be realized by real-time monitoring the key physiological parameters.The pulse oximeter,which is based on the photoplethysmography(PPG)technology,is a kind of contact measuring device.Due to its simpleness and accuracy,it has been used widely in clinical application.However,there are some unavoidable factors that greatly limit the use of the platform and even lead to inaccurate estimation during contact measurement.First of all,long time clamping between skin and sensors is needed,which affects people's daily life and habits.Besides,inaccurate measurement results will appear when the contact force and position is unsuitable.Further,contact PPG requires additional light sources,which greatly suppress detector flexibility.Therefore,a non-contact measurement method is needed to overcome these problems above.Here,imaging photoplethysmography(IPPG),which is derived from the PPG technology,is proposed.It is a non-contact,simple and low cost physiological information detection technology.This technology has been proposed as one of the hot spots in the field of Biomedical Photonics and medical instruments.Although some important physiological signals have been measured by IPPG technology,there are still some important problems such as motion artifact,light intensity and capture distance need to be solved.Thus,this paper aims to design three kinds of cardiovascular physiological parameters detection systems that based on the biomedical photonics,to achieve the physiological metabolic situation accurate monitoring.The accuracy and robustness of these detection systems are evaluated by multi-group experiments.Monochrome CMOS camera:A monochrome CMOS camera,dual-wavelengths LED lights and the corresponding band-pass filters are used.Arm is selected as the region of interesting(ROI)to analyze.The noise of the original signal and the motion artifacts that generated during the shooting process are removed by image segmentation and registration.Then,an adaptive filtering that based on wavelet decomposition is processed.Further,two-dimensional and three-dimensional distribution maps of blood oxygen saturation are constructed.The platform,which has high accuracy of oxygen saturation measurement,is suitable for clinical applications.Color CCD camera:A color CCD camera is utilized to calculate the pulse rate,oxygen saturation and heart rate variability under ambient light.The face region is selected as the ROI.In order to overcome the influence of motion artifacts,face detection,tracking and skin color detection are performed.Then blind source separation algorithm and wavelet decomposition are used to remove motion artifacts.A two-dimensional distribution colormap of the peak frequency of IPPG signal is proposed.Based on this colormap,the location and size of the ROI can be obtained.The measurement platform of the system is accurate.It is insensitive to changes of shooting conditions such as illumination and distance,and is suitable for clinical use and home healthcare.Webcam under ambient light:On the basis of the above research,in order to further promote the non-contact measurement,we analyzed the feasibility of monitoring the cardiovascular physiological parameters based on the consumption level camera.But,the video that gets from webcam has poor signal to noise ratio.At first,Euler video amplification is operated to make the weak signal to enlarge.An improved blind source separation algorithm based on principal component analysis and independent component analysis is proposed.The experimental results show that the non-contact measurement platform can obtain accurate measurement results remotely.Because of its hardware platform is easy to obtain,it has a huge application prospects in family health monitoring.The proposed non-contact systems cover the main IPPG scheme and try to effectively overcome the key factors such as motion artifacts,illumination and distance.Through the experiments of multiple volunteers,the results show that the three non-contact cardiovascular physiological parameters detection system can achieve the purpose of accurate prediction.There is a huge clinical and daily family health monitoring application.