Aperiodic Component Analysis in Quantification of Steady-State Visually Evoked Potentials.

Area of Science:

• Neuroscience
• Biomedical Engineering
• Signal Processing

Background:

• Steady-state visually evoked potentials (SSVEPs) are crucial for brain-computer interfaces (BCIs).
• The electroencephalogram (EEG) signal contains both periodic and aperiodic components.
• The influence of the aperiodic component on SSVEP quantification and BCI performance is not fully understood.

Purpose of the Study:

• To investigate the impact of the aperiodic component in EEG on SSVEP quantification.
• To determine how the aperiodic component affects the performance of different BCI algorithms.
• To compare the effectiveness of various analysis methods when controlling for the aperiodic component.

Main Methods:

• Applied the Fitting Oscillations & One-Over-F method to separate periodic and aperiodic EEG components.
• Measured EEG responses and BCI performance using four methods: power spectral density analysis (PSD), canonical correlation analysis (CCA), filter bank canonical correlation analysis (FBCCA), and task discriminant component analysis (TDCA).
• Compared method performance with and without controlling for the aperiodic component.

Main Results:

• Controlling for the aperiodic component significantly reduced BCI performance for CCA (94.9% to 82.8%), FBCCA (94.1% to 87.6%), and TDCA (96.5% to 70.3%).
• The aperiodic component had minimal impact on PSD analysis performance (80.4% to 78.7%).
• A differential impact of the aperiodic component was observed between PSD and the other three methods in differentiating target and non-target stimuli.

Conclusions:

• The aperiodic component significantly influences SSVEP quantification and BCI performance.
• The impact varies depending on the analysis method used.
• Future research on SSVEP quantification should consider the dynamic nature of aperiodic EEG activity.