The Accuracy of the Eye Tracking in a Virtual Reality Headset for Possible Research and Clinical Applications
Abstract
Virtual Reality (VR) headsets are widely used in various clinical and research settings. The reliability and quality of this technology heavily rely on the accuracy of data collected by the VR headsets. The specifications of the VR headsets' accuracy are normally published by the manufacturers. However, the performance claimed by the manufacturers is rarely validated by third-party organizations. Despite its importance, limited research has been focused on the data accuracy of VR headsets, even less so in eye-tracking applications. The main purpose of this project is to invest in the accuracy of eye rotation measurements recorded by the eye-tracking system built into a VR headset. A VR headset with eye-tracking capabilities (HTC Vive™ Pro Eye) was used in this study. For testing purposes, we also developed an eye model that can simulate human eye motion on a controlled pattern with a 30-degree range. The model was used to test the eye-tracking system data collection at frequencies ranging from 10 Hz to 500 Hz, with a step size of 10 Hz. A Unity program was developed to read and export the data, from which the headset’s eye tracking accuracy was assessed at each of the tested frequencies. Our experimental results suggested that the VR headset showed great potential for precise eye rotation measurement. Overall, the correlation between the VR headset measurement and the truth reference was between 0.97 and 0.99 from 10 Hz to 500 Hz. The root mean square error (RMSE) values were from 4.39 to 4.74 for the left eye and 3.63 to 3.67 for the right eye, both in degrees. We suspect that the increased RMSE values in the right eye may be due to the relative position between the VR headset and the eye system during testing. Nevertheless, the high correlation between the measurement and truth reference indicates precision in the estimation.
Keywords: Virtual Reality, eye tracking, HTC Vive Pro Eye, accuracy under different frequencies
DOI: 10.54941/ahfe1006185
Cite this paper
More from this volume
- Human Factors in an Agile Environment: Capturing Value in Healthcare
- Enhanced Fall Prevention in Nursing Facilities: Assisting Caregivers through Data-Driven Selective Monitoring and Notification
- Examining the Efficacy of an Improved CanSim for Quantifying Hemodialysis Cannulation Skills
- Human Factors Engineering and User- centered Design Principles in the Design and Development of Device Combination Products for Special Patients Populations
- Patients over Process: Stratifying Risk in the Design, Development, and Deployment of Artificial Intelligence in Healthcare
- Using Design Thinking to Improve Student Feedback in Healthcare Simulation
- Elicitation of risk perception strategies in emergency rooms based on KYT technique and eye tracking stimulated retrospections
- Advancing Scoliosis Treatment: Development and Evaluation of Anisotropic Textile Brace (ATB) for Enhanced Patient Compliance
- Biological evaluation of antimicrobial treated textiles
- Ergonomic investigation on Interventional Radiology in the era of robotic surgery
- Self-monitoring of blood glucose: The perception of physicians who care for older adults in a health service in Mexico
- Socio-Technical Risk Analysis for the Digitalized Transfusion Process: the e-TRAST Tool


AHFE Open Access