Logo image
Back
Implementation of a Home-Use Virtual Environment BCI for People with ALS Using Face Stimuli
Conference paper   Peer reviewed

Implementation of a Home-Use Virtual Environment BCI for People with ALS Using Face Stimuli

Emma Elizabeth Dryden, Terry D Heiman-Patterson and Hasan Ayaz
Muscle & Nerve, v 66(S2), pp S33-S34
01 Nov 2022
DOI: https://doi.org/10.1002/mus.27729 (73)

Abstract

Clinical Neurology Life Sciences & Biomedicine Neurosciences Neurosciences & Neurology Science & Technology
Brain-Computer Interfaces (BCI) can be used as alternative communication for individuals with Amyotrophic Lateral Sclerosis (ALS). P300 BCIs predict where the user is looking based on event-related potentials in response to flashing stimuli. Despite the ability of P300 BCIs to enable communication for people with severe physical disability, home based use of BCIs is still not part of the care continuum. In part, this is likely due to the progression and physical challenges of ALS making the engagement with BCI set-ups difficult. In an effort to improve the usability of BCIs in the home for people with ALS, we tested whether presenting visual targets in the form of faces would improve accuracy. Previous studies have shown that BCI performance improves in controls when using faces, including both familiar and non-familiar faces, as the visual stimulus. 16 individuals volunteered (8 ALS and 8 controls). Six different types of face stimuli were implemented: Blank (for control), male/female celebrity and stranger faces, and familiar (close family) face. The study was conducted at the participant’s home with a miniaturized BCI system, including a wireless 8-channel mobile EEG (Unicorn). A virtual navigation environment was created with movement commands integrated from the BCI to increase user engagement. Online BCI accuracy was lower for ALS users (45.6%) versus healthy controls (71.7%). A significant correlation was found between BCI online accuracy and ALSFRS-R score (R2 =0.725), supporting our previous findings. Visually evoked potential (VEP) latency was similar between ALS and HC groups, with decreased amplitude in the ALS group. The face-specific N170 component was greatest in amplitude in the Familiar condition, and smallest in the Blank condition. Lower ALS accuracy could be due to physical discomfort and fatigue due to disease. Results suggest the possibility of better BCI performance because of the robust early VEP associated with the sociocognitive role in facial processing, verified in similar studies. This miniaturized BCI is usable for both ALS and control subjects and using face-specific visually evoked potentials could further personalization of BCIs. Future work includes alternative signal processing and hybrid approaches to further improve BCI use with ALS.

Metrics

49 Record Views

Details

InCites Highlights

Data related to this publication, from InCites Benchmarking & Analytics tool:

Web of Science research areas
Clinical Neurology
Neurosciences
Logo image