Brain-computer interface and semantic classical conditioning of communication in paralysis

Area of Science:

• Neuroscience
• Biomedical Engineering
• Communication Science

Background:

• Brain-computer interfaces (BCIs) commonly use instrumental-operant learning for communication in locked-in states.
• This approach requires voluntary motor control, which is absent in the completely locked-in state.

Purpose of the Study:

• To propose and evaluate a classical semantic conditioning procedure as an alternative communication method for the completely locked-in state.
• To establish cortical responses to the truthfulness of statements, independent of specific words or sounds.

Main Methods:

• Acoustic presentation of true and false statements.
• Use of brief electrical pulses as unconditioned stimuli, paired only with true statements.
• Electroencephalographic (EEG) recordings from 15 healthy participants and one ALS patient.
• Application of three distinct classifiers to differentiate cortical responses.

Main Results:

• Successful differentiation between cortical responses to true and false statements.
• Demonstration of semantic classical conditioning's efficacy in enabling communication.
• Validation of a non-muscular communication channel for locked-in individuals.

Conclusions:

• Classical semantic conditioning offers a viable alternative for basic communication in the completely locked-in state.
• This method bypasses the need for instrumental-operant learning and motor control.
• EEG-based analysis of conditioned cortical responses can facilitate communication for severely paralyzed individuals.