Kinect Modelling of Chest Compressions - A Feasibility Study for Chest Compression Depth Measurement Using Digital Strategies

Quality cardiopulmonary resuscitation (CPR) increases the chances of survival from out-of-hospital cardiac arrest. CPR measurement devices with real-time feedback could assist in the provision of this. Others have proposed accelerometer-based feedback systems by using specialized cards, smartwatches or hand-held smartphones. Our group have previous proposed a system that measure chest compression (CC) rate and hands-off-time utilizing a smartphone camera with a phone-on-the-floor solution. In this paper we have investigated the possibilities of also measuring the important CPR metric CC depth. Solutions using smartwatches or smartphones estimate CC parameters based on the bystanders movement. However, there are no reported work on analyzing different bystanders movement during CCs. In this work, a CC modelling experiment using Microsoft Kinect is performed to measure the degree of variations in CC techniques, providing knowledge on limitations when considering digital strategies for CC depth measurements. Although variations between the CC techniques were discovered, the results indicate that smartphone depth-cameras and accelerometer-sensors could in most cases be used for CC depth measurement with acceptable accuracy.