Brain-computer communication: self-regulation of slow cortical potentials for verbal communication

Abstract

Objective: To test a training procedure designed to enable severely paralyzed patients to communicate by means of self-regulation of slow cortical potentials.

Design: Application of the Thought Translation Device to evaluate the procedure in patients with late-stage amyotrophic lateral sclerosis (ALS).

Setting: Training sessions in the patients' homes.

Participants: Two male patients with late-stage ALS.

Interventions: Patients learned voluntary control of their slow cortical potentials by means of an interface between the brain and a computer. Training was based on visual feedback of slow cortical potentials shifts and operant learning principles. The learning process was divided into small steps of increasing difficulty.

Main outcome measures: Accuracy of self-control of slow cortical potentials (percentage of correct responses). Learning progress calculated as a function of training session.

Results: Within 3 to 8 weeks, both patients learned to self-regulate their slow cortical potentials and to use this skill to select letters or words in the Language Support Program.

Conclusions: This training schedule is the first to enable severely paralyzed patients to communicate without any voluntary muscle control by using self-regulation of an electroencephalogram potential only. The protocol could be a model for training patients in other brain-computer interface techniques.