Detection of chest compressions in resusciatation data

Abstract

Cardiac arrest is the most common reason of death for thousands of people every year. Immediate treatment of cardiac arrest can reduce the rate of death. The treatment of the cardiac arrest involves cardiopulmonary resuscitation (CPR) providing chest compressions and ventilations, medicine and defibrillator to analyse the heart rhythm and give an electric shock if the rhythm is shockable, this kind of defibrillator called automated external defibrillator (AED). A good quality of CPR can increase the survival rate, but in practice the quality of CPR falls rapidly after short time and not keeping the recommendation of guidelines for CPR. In order to improve the quality of CPR, it has been suggested to improve the automated external defibrillators (AED) with the ability to check the quality of CPR and give a feedback if it falls. To achieve this improvement it is important to know when the CPR is ongoing, in other words it will be very useful to detect the chest compressions and ventilations in the heart rhythm such that the AED can analyse the information of the detection to give a feedback to improve the CPR when it is needed. In this study we will try develop a detection system of the chest compressions. A method to annotate the compressions will be developed also to make the reference (annotations) of the chest compressions that allow us to evaluate the performance of the detection system. The pressure information which is measured by sensor on an extra pad of the defibrillator is used for the annotation of the chest compressions and the thoracic impedance signal is used to develop the detection system. Performance of the detection system based pressure/Impedance/ECG signal is evaluated and presented in this thesis.