Identifying P300 Brain-Computer Interface Training Strategies for AAC in Children: A Focus Group Study

Authors

Kevin M. Pitt, University of Nebraska-Lincoln
Jamie B. Boster, Nationwide Children's Hospital, Columbus, Ohio

Document Type

Article

Date of this Version

2025

Citation

Augmentative and Alternative Communication (2025)

doi: 10.1080/07434618.2025.2495912

Comments

Used by permission

Abstract

The integration of Brain-Computer Interface (BCI) technology into Augmentative and Alternative Communication (AAC) systems introduces new complexities in training, particularly for children with diverse cognitive, sensory, motor, and linguistic abilities. Effective AAC training is crucial for enabling individuals to achieve personal goals and enhance social participation. This study aimed to explore potential training strategies for children using P300-based BCI-AAC systems through focus group discussions with experts in AAC and BCI technologies. Participants identified six key themes for effective training: (1) Scaffolding—developing adaptive systems tailored to each child’s developmental level, including pre-teaching, visual display adaptations, and gamification; (2) Verbal Instructions—emphasizing the use of clear, simple language and spoken prompts; (3) Feedback—incorporating immediate feedback and biofeedback methods to reinforce learning; (4) Positioning—ensuring proper trunk stability and addressing electrode placement; (5) Modeling and Physical Supports—using physical cues and demonstrating BCI-AAC use; and (6) Considerations for Visual Impairment—accommodating cortical visual impairment (CVI) with suitable stimuli and environmental adjustments. These insights provide an initial foundation for identifying BCI-AAC training strategies for P300-BCI-AAC use by children. Further systematic research involving end users and their support networks, along with other professions, is needed to validate, enhance, and identify new intervention approaches that support communication outcomes for children with varying needs.