Agentic EEG Interfaces for Screen-Independent Assistive Communication in ALS - 2026_CDR_DEIB_13

This project aims to develop a non-invasive, screen-independent EEG communication system for people with ALS who struggle with speech, movement, or eye-tracking tools. It will detect intentional attempts to communicate through imagined words or short phrases and identify messages from a personalized vocabulary of around 20 essential expressions. A constrained AI assistant will manage uncertainty, request confirmation, offer alternatives, and convey confirmed messages while keeping users in control. It will support limited, intentional communication, not general thought reading. Without developing new hardware, the project will test fewer EEG channels, shorter calibration, variable signal quality, and repeated use. Expected outputs include a research prototype, validated benchmarks, and requirements for future testing with ALS patients and wearable implementation.