Unconstrained Detection Of Ballistocardiogram And Heart Rate Based On Vibration Acceleration

In recent years,with the change of people’s eating habits and the aging of the population,the incidence of cardiovascular disease is increasing.The key to the prevention of cardiovascular disease is to carry out a scientific heart monitoring in daily life,so as to discover the characteristics of the disease in time.But conventional monitoring methods,such as electrocardiogram,echocardiography,magnetograph and so on,all need professional equipment,which make users constrained,and difficult to be used for daily monitoring.However,the accuracy of unconstrained monitoring devices like health bracelet and intelligent pillow was unable to meet the clinical requirements.During the process of pumping blood,the heart beat will cause stress changes or mechanical vibration of the support closely contacting with the human body.The Ballistocardiogram(BCG)signal can be obtained from these vibration data by means of electrical measurement.The BCG signal contains rich cardiac information,which can carry out non-sensory measurement,and even long-term monitoring will not bring the body and mental burden to the subjects.Based on these advantages,a method was proposed to detect BCG information and then extract heart rate from the vibration signal of the bed.First of all,this paper analyzed the needs of monitoring environment and users,and determined that the detection mode is sleep monitoring.In view of the shortcomings of traditional research,a method using 3 axis accelerometer to collect the vibration data of the bed excited by the impact and detect the BCG signal from it was proposed.At the same time,through theoretical analysis and experimental modal analysis,the best direction of detecting BCG signals in sleep state was found,that was the direction of the spine.Then,by comparing the existing signal preprocessing method,the Butterworth filter was chosen to denoise the signal,and the parameters of the filter were determined through simulation experiment.Based on the theory of signal similarity,this paper compared the segmentation gradient filtering results of the original signal with the theoretical BCG signal,then found the best filter band.During the extracting of heart rate,this paper developed a J wave detection method based on BCG energy waveform,and proposed a method of heart rate calculation combined with adaptive amplitude threshold and periodic threshold,avoided the wrong detection phenomenon of traditional detection methods.Finally,the individual difference,robustness and applicability of the detection system were verified.The experimental results showed that the detection system had high accuracy for the different experimental individuals.It could handle the distorted BCG signal with weak interference and exclude the influence of strong interference.And it was suitable for the comfortable environment of daily life.This had laid a solid experimental foundation for the clinical application of the unfettered heart rate monitoring technology.