Related Experiment Video
Updated: Jul 10, 2026
07:37
Assessment and Communication for People with Disorders of Consciousness
Published on: August 1, 2017
Classification of imagined beats for use in a brain computer interface
Summary
Brain-computer interfaces can use imagined rhythms for information transfer. Researchers developed a system classifying imagined musical rhythms from electroencephalography (EEG) signals, achieving over 0.7 accuracy with 2 seconds of data.
Area of Science:
- Neuroscience
- Biomedical Engineering
- Signal Processing
Background:
- Electroencephalography (EEG) signals exhibit power spectrum changes when anticipating auditory stimuli.
- These EEG changes persist even without actual auditory input, suggesting potential for brain-computer interfaces (BCIs).
Purpose of the Study:
- To investigate the feasibility of using imagined rhythms for information transfer via BCIs.
- To develop and evaluate a classifier distinguishing between imagined accented and non-accented tones using EEG signals.
Main Methods:
- Four healthy subjects imagined simple rhythms, guided by a metronome.
- EEG signals were analyzed, focusing on phase and power differences of independent components.
- Features for classification were automatically selected from a predefined set.
Main Results:
- Classification accuracy reached approximately 0.6 for two subjects.
- Combining classifications improved accuracy to over 0.7 for the best-performing subject using 2 seconds of data.
- The classification task's chance level was 0.5.
Conclusions:
- Imagined rhythms can be reliably detected from EEG signals.
- This demonstrates the potential for developing BCIs based on imagined auditory patterns.
- Further development could enhance classification rates for practical BCI applications.
