Enhancing SSVEP-BCI Performance Under Fatigue State Using Dynamic Stopping Strategy

Abstract

Steady-state visual evoked potential (SSVEP)based brain-computer interfaces (BCIs) have gained significant attention in recent years due to their advantages, including high communication rates, short calibration times, and high signal-to-noise ratios. However, one of the major challenges in practical application is performance degradation due to fatigue during continuous and prolonged use. This study aims to mitigate the negative effects of fatigue on SSVEP-BCI performance by adjusting data lengths and updating detection models using dynamic stopping strategy. Two 16-target SSVEP-BCIs were used for data collection, one using low-frequency stimulation and the other using high-frequency stimulation. A self-recorded fatigue dataset from twenty-four subjects was used for extensive evaluation studies. The simulated online experiment demonstrated that the proposed methods outperformed the conventional fixed stopping strategy in terms of classification accuracy, information transfer rate, and selection time, regardless of stimulus frequency. In conclusion, the proposed methods can improve the performance of the SSVEP-BCI in fatigue state and provide superior performance over extended periods of sustained use.