Aphasia recovery by language training using a brain-computer interface: a proof-of-concept study

Area of Science:

• Neuroscience
• Rehabilitation Medicine
• Biomedical Engineering

Background:

• Aphasia, a common post-stroke language disorder, has limited recovery with conventional therapy.
• Current speech and language therapy show small effect sizes in chronic stroke patients.
• Brain plasticity offers a potential target for enhancing aphasia rehabilitation.

Purpose of the Study:

• To introduce and evaluate a novel brain-computer interface (BCI) based language training for aphasia.
• To investigate the feasibility and efficacy of high-intensity BCI training in chronic aphasia.
• To explore the neural mechanisms underlying BCI-induced language recovery.

Main Methods:

• A BCI system recorded EEG signals during an auditory target word detection task.
• Machine learning models provided real-time, brain-state-dependent feedback to patients.
• A proof-of-concept study involved 10 chronic aphasia patients undergoing 30 hours of training.

Main Results:

• High-intensity BCI training was feasible, even with rapid word presentation and challenging EEG signals.
• Significant medium to large improvements were observed in multiple language assessments post-training.
• Five patients transitioned from aphasic to non-aphasic status, indicating substantial recovery.

Conclusions:

• BCI-based language training offers a promising new avenue for aphasia rehabilitation.
• The training induced sustained language recovery without affecting attention or non-linguistic skills.
• Neuroimaging suggests BCI training strengthens language networks and promotes neural rebalancing.