Related Experiment Video
Updated: May 2, 2026
10:51
An Experimental Platform to Study the Closed-loop Performance of Brain-machine Interfaces
Published on: March 10, 2011
13.7K
Learning to operate an imagined speech Brain-Computer Interface involves the spatial and frequency tuning of neural
Kinkini Bhadra1, Anne-Lise Giraud1,2, Silvia Marchesotti3
1Department of Basic Neurosciences, Faculty of Medicine, University of Geneva, Geneva, Switzerland.
Communications Biology
|February 20, 2025
Summary
Learning to control brain-computer interfaces (BCI) for communication improves with practice. Continuous feedback is essential for this learning, which involves dynamic brain activity changes, enhancing BCI usability.
Area of Science:
- Neuroscience
- Biomedical Engineering
- Rehabilitation Technology
Background:
- Brain-Computer Interfaces (BCI) offer communication potential for individuals with speech impairments.
- Current BCI research emphasizes signal decoding over user adaptation.
- Understanding user learning dynamics is crucial for BCI improvement.
Purpose of the Study:
- To investigate if BCI control improves with user training.
- To characterize the neural dynamics associated with BCI learning.
- To determine the role of feedback in BCI skill acquisition.
Main Methods:
- 15 healthy participants trained on an electroencephalography (EEG)-based BCI for imagined speech.
- Five consecutive days of training with syllable imagery.
- Control experiment to assess the necessity of continuous feedback.
Main Results:
- Significant global improvement in BCI control observed across participants.
- Continuous feedback was demonstrated as necessary for learning.
- Learning correlated with increased frontal theta and temporal gamma EEG activity.
Conclusions:
- User training significantly enhances BCI controllability.
- BCI learning involves adaptive neural changes across different frequency bands.
- Combining machine and human learning optimizes BCI performance.
