A Virtual Reality System For Realistic Cardiopulmonary Resuscitation Training

Abstract

Introduction: In the United States, more than 350,000 out-of-hospital cardiac arrests occur annually. CPR, while lifesaving, needs to be performed accurately to be effective. This study aims to implement virtual reality (VR) with wearable sensors to measure compression acceleration, frequency, and depth while providing real-time feedback to the user for immersive and effective CPR training. Materials and Methods: VR scenarios were customized in Unity video game engine to serve as the user interface and feedback environment. An HTC Vive VR system immersed the user into a real-world CPR scenario while an HTC Vive tracker tracked the position of a CPR mannequin in VR. An MPU6050 accelerometer was programmed with an Arduino Micro to process compression acceleration data. Leap motion hand tracking technology was utilized to track the user’s hand position. Speech recognition was implemented to recognize key phrases, such as “call 911,” spoken by the user to a non-playable character in the VR scenario. Quantitative and qualitative performance feedback was provided to the user in the form of visual and auditory cues to correct CPR compressions. Conclusions: This system is an innovative proof of concept for a VR-based CPR training system that can provide compression feedback for high quality CPR.