The use of EEG and VR in Mental State Recognition

Abstract

Mental state recognition (MSR) is important to multiple health-related fields. A virtual reality (VR) headset is used to induce mental states through both distractors and stressors.Participants solved arithmetic questions in VR, then outside of VR using a Muse S EEG device. A heart monitor was used throughout. Three research contributions followed: First, heart rate variability (HRV) data were compared between VR and non-VR sessions and correlated with established Test of Variables of Attention (T.O.V.A.) measurements used to asses participants' attention and focus. Second, a classifier was developed to differentiate between clean and noisy EEG data, with 92% accuracy. Lastly, linear regression models were developed, achieving mean squared error scores of 0.65 and 0.63 for 3-level stress and attention prediction from EEG data, respectively. In summary, this thesis explores the use of VR to induce mental states and advances the state of the art in EEG-based MSR.