A quantization algorithm of visual fatigue based on underdamped second order stochastic resonance for steady state...

Area of Science:

• Neuroscience
• Biomedical Engineering
• Signal Processing

Background:

• Visual fatigue from steady state visual evoked potential (SSVEP) paradigms impacts brain-computer interface (BCI) applications.
• Current objective quantification of SSVEP visual fatigue relies on canonical correlation analysis (CCA), which has limitations.
• Developing reliable methods to measure SSVEP-induced visual fatigue is crucial for advancing BCI technology.

Purpose of the Study:

• To introduce a novel algorithm for quantifying visual fatigue in SSVEP paradigms using single-channel EEG.
• To evaluate the proposed algorithm's accuracy and reliability against established methods and subjective measures.
• To enhance the practical application of BCIs by addressing the challenge of visual fatigue.

Main Methods:

• Developed a new visual fatigue quantification algorithm based on underdamped second-order stochastic resonance (USSR).
• Employed a fixed-step energy parameter optimization algorithm coupled with the USSR model to enhance signal-to-noise ratio.
• Compared the USSR-based algorithm with traditional CCA and the Likert fatigue scale for validation.

Main Results:

• The USSR-based algorithm showed no significant difference (p=0.090) when compared to the subjective Likert fatigue scale.
• Traditional CCA demonstrated a significant difference (p<0.001) compared to the subjective Likert fatigue scale.
• The proposed algorithm achieved higher agreement with the subjective gold standard for visual fatigue quantification.

Conclusions:

• The USSR-based algorithm provides a more reliable and accurate quantification of SSVEP-induced visual fatigue.
• This new method offers a superior alternative to CCA for objective visual fatigue assessment in SSVEP paradigms.
• The findings support the potential of the USSR algorithm to improve the user experience and effectiveness of BCIs.