An XR Experience to Collect Biosignals for Cybersickness Mitigation

Cybersickness (CS) is one of the main obstacles to the use of Virtual Reality (VR), often caused by Head-mounted Displays (HMDs). Its symptoms, which can last from minutes to hours after exposure, include nausea, vertigo, eye strain, and headache. Researchers generally use subjective measures, such as the Virtual Reality Sickness Questionnaire, to assess CS. Studies indicate that CS significantly impacts physiological signals, but there is little research on the application of Symbolic Machine Learning to understand the causes of CS in VR games. This study investigates the use of biosignals to identify the causes of CS in VR. Our main hypothesis is that the combination of quantitative and subjective assessments, along with Symbolic Machine Learning techniques, allows for the creation of a ranking of the main causative or indicative factors of CS. To validate this hypothesis, software was developed to record the biosignals and self-reported symptoms of participants during experiments with two VR games. Physiological signals (ECG, EDA, and body movements extracted from an Accelerometer - ACC) and game data were collected. The results show a strong relationship between physiological changes and CS symptoms, with a model that includes biosignals achieving an AUC of 0.95. The rankings of the main factors, both for the model without and with the inclusion of biosignals, confirmed previous research described in the literature. As far as we know, our work is the first to use Symbolic Machine Learning models to detect the causes of CS.