Research And Realization Of Intelligent Control System For Cardiopulmonary Resuscitation In Rats

Cardiopulmonary resuscitation(CPR)is the only effective way to rescue cardiac arrest patients.Chest compression is the main aspect of cardiopulmonary resuscitation.The quality of chest compression is of great significance to the success of first aid.However,compression frequency and depth which are recommend by 2010 American Heart Association(AHA)cardiopulmonary resuscitation and emergency cardiovascular care(ECC)guide are not yet popularly accepted.The improvement on the process of CPR becomes extremely urgent.At present,the animal experiment is often used in the improvement of CPR.As experimental objects,rats are usually chosen in CPR after cardiac arrest.Manual chest compression is widely used in the experiment of CPR in rats.The experimental operator adjusts the compression pattern according to the physiological index based on experience.Manual chest compression is difficult to achieve accurate,stable and effective chest compressions,which harms the development of the improvement on CPR.According to the issues discussed above,a chest compression equipment with automatic controlled and an intelligent control method are proposed based on machine learning classification algorithm for CPR in rat.The equipment would improve the successful rate of CPR in cardiac arrest rats effectively,and provide a theoretical guide for CPR.This method is of great significance in the improvement of CPR.The research contents of this paper are the following three parts.1.Chest compression equipment for CPR in rat was designed.Considering the functional requirements and performance indexes of the equipment,the threedimensional design is completed in the CAD software SolidWorks.The PLC is used as the slave computer to control the servo motor as the way to drive the compress bar.Reciprocal linear movement of compress bar is driven by servo motor with ball screw and linear guideway.In the.NET platform,the development of the host computer control software is completed with the C# language.Through the method,the control of compress operation and the display and storage of physiological indexes are realized by computer software.2.Characteristic parameters reflecting the physiological state during CPR in rats are collected and proceeded.Aorta blood pressure signal,temperature signal,and pressure signal are collected during CPR in rat.Six physiological characteristic parameters of systolic blood pressure(SBP),diastolic blood pressure(DBP),mean blood pressure(MBP),stroke volume(SV),K value and main wave rise slope(S)are proposed based on the mechanism of chest compressions.Then,18 characteristic parameters reflecting the individual characteristics and pressing characteristics during CPR in rats are summed up.These parameters reflect the physiological state of the rat during CPR.The algorithm to confirm boundary based on wavelet transform is introduced to obtain the blood pressure wave’s peaks.In particularly,the characteristic parameters are calculated and standardized.3.The intelligent control method for CPR in rats is proposed.The experimental process is transformed into a multi-dimensional feature set through transformation.The prediction model of spontaneous breath recovery in rats is established based on logistic regression algorithm,Bayes discriminant algorithm and C4.5 decision tree according to the machine learning classification algorithm.The key factors that influence the spontaneous breath recovery of cardiac arrest rats are proposed.Finally,based on the key factors,the intelligent control method for CPR in rats was designed and verified.