Design And Research Of Real-time Monitoring System For Chest Compressions

Cardiac arrest is one of the most deadly diseases that threaten human life.Most patients are out of the hospital at the time of onset due to the unpredictability.Therefore,patients’ survival rates can be greatly improved if given effective cardiopulmonary resuscitation(CPR)in time.Though the American Heart Association has given guidelines for the CPR process,it is still exceedingly difficult for first responders to meet CPR standards due to a lack of first aid experience.In order to ensure that emergency personnel can maintain correct and effective chest compressions during cardiopulmonary resuscitation,a set of real-time monitoring system for chest compressions is designed and developed,real-time monitoring of chest compressions and feedback of the quality of chest compressions are realized.The real-time chest compression monitoring system designed enjoys the advantages of compact,low cost,and portable.The system can be used for training or clinical assistance.In this paper,the author firstly studied the model of the chest compression CPR,according to its characteristics of multi-dimensional,multi variable and high complexity.And then,studied the data processing algorithm based on sensor real-time data acquisition.An improved empirical mode decomposition(EMD)algorithm is used for analyzing the experiment result.Additionally,the B-spline function is used to replace the cubic spline interpolation function of the empirical mode decomposition method.The usage of the B-spline function not only ensured the timeliness of chest compressions but also removed the acceleration trend in the time domain.The starting point and endpoint of compression could be found by analyzing the velocity waveform after the one-time integration of acceleration data.Furthermore,the calculation method proposed based on compression depth,chest compression frequency and whether the compression rebounds fully by using the second integration method to obtain the displacement signal of chest compression.The hardware design of chest compression is implemented using Bluetooth low energy chip and acceleration sensor,and a mobile phone application software is designed on the Android operating platform to realize the host computer processing of chest compression data.In addition,the method of feature extraction for CPR of chest compression is studied.An improved feature extraction method based on EEMD and multi-scale entropy is proposed.Support vector machine(SVM)was used to identify the standard motions of chest compression CPR.