Rishabh Chandel
Pronouns:
Research Mentor(s): Jane Huggins
Research Mentor School/College/Department: Physical Medicine and Rehabilitation / Medicine
Program: UROPF
Session: Session 5 (2:40pm – 3:30pm)
Authors: Rishabh Chandel, Jane E. Huggins
Abstract
Through extensive research brain-computer interfaces, or BCIs, have greatly improved during the past few decades. As described by Farwell and Donchin (1988), these are devices to provide communication to individuals who often don’t have any motor control. However, there still exists a major hurdle when calibrating BCIs to establish communication with impaired individuals having low attention spans, as active response to stimuli is often inaccurate and inconsistent, as discussed by Huggins, Karlsson, and Warschausky (2022). To combat this, Dr. Huggins has founded the choice-making project, which redesigns the calibration process. Unconscious reactions to rare stimuli and their more common, scrambled counterparts are both measured, which ideally allows the BCI to separate EEG generated by both stimulus types and use the resulting weights to predict future user choices between stimuli. The calibration process is then tested by presenting the user with two stimuli to choose from: one high arousal and one low arousal image. The BCI uses the results to calculate what stimulus the user unconsciously pays attention to, and this choice is either confirmed or denied by the user. In order to realize these goals, the current solution was first debugged, bringing to light and fixing conflicts between different modules that have been propagated through recent git commits. Parameter files were created to test the proposed calibration process, as well as to compare side-by-side vs one-by-one image presentation types, to obtain preliminary results. In order to promote longer attention spans, gamification was implemented through rewarding users with videos for correctly focusing on target images. This work will allow for further testing of the choice-making BCI to produce more extensive results and determine which presentation type yields more accuracy. With these steps, the user should be able to follow instructions and focus on an image between two stimuli with equal variance, with the BCI accurately identifying what image is selected.
