Related Experiment Video
Updated: Jul 9, 2025
06:34
A Single-Channel and Non-Invasive Wearable Brain-Computer Interface for Industry and Healthcare
Published on: July 7, 2023
2.4K
Using oscillatory and aperiodic neural activity features for identifying idle state in SSVEP-based BCIs reduces false
Rui Wang1, Tianyi Zhou2, Zheng Li2,3
1Department of Electrical Engineering and the Key Laboratory of Intelligent Rehabilitation and Neuromodulation of Hebei Province, Yanshan University, Qinhuangdao 066004, People's Republic of China.
Journal of Neural Engineering
|November 28, 2023
Summary
This study introduces a new method for brain-computer interfaces (BCIs) that combines brain signal rhythms and non-rhythmic activity. This fusion improves the accuracy of identifying idle states, reducing false triggers in asynchronous BCIs.
Area of Science:
- Neuroscience
- Biomedical Engineering
- Signal Processing
Background:
- Traditional brain-computer interface (BCI) methods primarily use rhythmic (oscillatory) brain activity, neglecting aperiodic (1/f) components.
- Recent research suggests aperiodic activity is linked to cognitive functions, but its role in distinguishing brain states for BCIs remains unclear.
- Asynchronous BCIs require robust methods to differentiate between control and idle states, minimizing false triggers.
Purpose of the Study:
- To develop and evaluate an asynchronous BCI method that fuses oscillatory and aperiodic features for improved steady-state visual evoked potential (SSVEP) based classification.
- To determine if aperiodic brain activity can enhance the accuracy of idle state recognition in BCIs.
- To reduce false triggers in asynchronous BCIs by leveraging both rhythmic and non-rhythmic neural signal components.
Main Methods:
- Implemented irregular-resampling auto-spectral analysis to evaluate oscillatory and aperiodic components of brain states.
- Extracted oscillatory features (spectral power of fundamental, second, and third harmonics) and aperiodic features (slope and intercept of spectral fit).
- Applied feature selection using Bonferroni corrected p-values from ANOVA and utilized statistically significant, spatial-specific features for classification.
Main Results:
- The fused oscillatory and aperiodic features achieved an average accuracy of 88.39% in idle state recognition.
- This represents a significant improvement of 4.86% compared to using only oscillatory features (83.53% accuracy).
- The method demonstrated enhanced performance in distinguishing idle states by incorporating aperiodic brain activity features.
Conclusions:
- Aperiodic brain activity features are effective for recognizing idle states in asynchronous BCIs.
- Fusing oscillatory and aperiodic features significantly enhances classification performance compared to using oscillatory features alone.
- The proposed method offers a more robust approach for idle state detection, potentially reducing false triggers in SSVEP-based BCIs.
