| With the improvement of living standards and the deepening of aging,the prevalence of cardiovascular disease is on the rise,which seriously affects people’s health.Intelligent monitoring equipment and wearable devices provide convenience for people’s daily physiological signal monitoring,and provide important clinical reference value for disease diagnosis and prevention.The current monitoring equipment for human pulse,heart rate and blood oxygen signals such as Smart Wristband and Smart Watches are based on photoplethysmography technology(PPG),but PPG sensor is essentially a flow sensor,the pulse wave of PPG is not the same as the pulse wave taken by taking pulse and the real pulse wave features cannot be restored by PPG sensor.the signal is easily disturbed by the background light,resulting in the loss of the pulse wave features.Compared with the PPG sensor,the signal collected by the pressure sensor is closer to the essence of the pulse signal,has higher detection sensitivity and stability,and can obtain more detailed features of the pulse wave.Therefore,in this paper we will carry out research on pulse signal monitoring technology based on curved optical fiber pulse pressure sensor,and use the features extracted from pulse wave for prediction of blood pressure(BP)and pulse wave transit time(PTT).In this paper,we firstly introduce the causes and significance of physiological parameters such as pulse wave,blood pressure and pulse wave transmit velocity,and introduce the working principles,characteristics and limitations of different pulse wave monitoring technologies.Then we propose a pulse wave signal acquisition system based on curved optical fiber pulse pressure sensor,the results show that the system can restore the pulse signal well.Then,in order to realize the prediction of BP and PTT by pulse wave features,we use the Support Vector Regression(SVR)optimized by Genetic Algorithm(GA-SVR).for obtaining the data set required for training,the curved optical fiber pulse sensing system is used to obtain the waveform of the radial artery pulse wave and extract the pulse wave features,the electrocardiogram(ECG)sensor and PPG sensor are used to obtain the PTT,and the BP is recorded by the sphygmomanometer.During the experiment,the physiological data of 150 volunteers were collected,the establishment of the data set was completed,and the data set was randomly divided into a training set and a test set according to the ratio of 13:2.and then the Genetic Algorithm was used to find the penalty factor and kernel function parameters.Input the training set data into the GA-SVR model for training,and use the test set data to verify the accuracy of the model.The results show that BP and PTT can be predicted by the pulse wave features via GA-SVR model.The prediction accuracy is 91.48%,92.13%,87.49% for SBP,DBP and PTT,respectively.Finally,so as to improve the practicability of the system,we design a smartphone APP for displaying the results on the HBuilder-X software platform.For the convenience of users,we also design a personal information input window and pulse rate,Body Mass Index(BMI),BP signal display window. |
