Developing an Auditory Hybrid Brain-computer Interface Based on Near-infrared Spectroscopy and Electroencephalography
Year: 2016
Type: dissertation
Abstract: Auditory brain-computer interfaces (BCIs) serve as an access pathway for individuals who present as locked-in and have concomitant visual impairments, but they do not perform as well as visual BCIs. Since hybrid BCIs that combine electroencephalography (EEG) and near-infrared spectroscopy (NIRS) outperform unimodal BCIs in motor imagery tasks, we hypothesized that such a hybrid BCI would outperform a unimodal BCI in an auditory oddball task. Eleven non-disabled adults (aged 25.6 Âą 3.3) participated in two offline sessions and eight returned for a third online session. While combined EEG-NIRS classification did not outperform EEG-exclusive classification, combined classification does show modest gains in 2 participants when using all the NIRS data but constraining the EEG data to only a single oddball. In conclusion, NIRS offers limited benefits in terms of classification in a prolonged auditory task paradigm, but future research should focus on optimizing task duration for EEG and NIRS classification.
Source: TSpace (University of Toronto)
Author: Stefanie Glowinsky
Language: English
Subfield: Cognitive Neuroscience
Field: Neuroscience
Domain: Life Sciences
Open Access status: green
Funder: Bloorview Research Institute
OpenAlex ID: W3088517775