UC Santa Barbara

The recent proliferation of VR and AR devices has led to an increase in the use of these devices as research tools. As these technical developments continue, researchers can leverage these hardware improvements to create realistic and controlled environments for experimentation in life-like scenarios. In cognitive research, these devices will often be coupled with neurophysiological recordings, which poses the challenge of dealing with movement artifacts. In this study, three experiments were conducted using oddball tasks and semantic processing tasks to assess EEG data quality using VR/AR to display stim- uli. The first experiment showed that the VR oddball task elicited comparable neural activity as a traditional desktop oddball task. The subsequent experiments systematically introduced movement artifacts in VR and AR, and showed that these neural data were usable with minor movement artifacts, while neural signals recorded under walking/free motion conditions were heavily contaminated with movement artifacts. Although there have been a variety of approaches for removing movement artifacts from neural data, many of them are specific to the experimental design, or have other constraints limiting their generalizability.