Implementation of a Smart Teaching and Assessment System for High-Quality Cardiopulmonary Resuscitation

Abstract

The purpose of this study is to develop a smart training and assessment system called SmartCPR, for teaching and training cardiopulmonary resuscitation (CPR), based on human posture estimation techniques. In this system, trainees can automatically recognize and evaluate whether chest compressions during CPR meet the standard of high-quality CPR by simply using a device such as a smart phone. Through the system, trainees are able to obtain real-time feedback on the quality of compressions so that they can adjust the cycle, depth, frequency, and posture of compressions to meet the standard of high-quality CPR. In addition, the SmartCPR system is convenient for CPR trainers. Trainers can instantly and accurately assess whether the trainee’s compressions meet the standard of high-quality CPR, which reduces the risk of manual assessment errors and also reduces the trainer’s teaching pressures. Therefore, the SmartCPR system developed in this study can be an important tool for CPR teaching and training for physicians, which can provide training and guidance for high-quality CPR maneuvers and enable trainees to become more proficient in CPR and self-training.

Keywords

• first aid training
• cardiopulmonary resuscitation
• human pose estimation
• deep learning