Estimation of overlapped Eye Fixation Related Potentials: The General Linear Model, a more flexible framework than...

Area of Science:

• Neuroscience
• Cognitive Science
• Biomedical Engineering

Background:

• Eye Fixation Related Potential (EFRP) estimation relies on averaging EEG signals at ocular fixation onset.
• A key limitation is signal overlap due to short Inter Fixation Intervals (IFI), where multiple fixations within an epoch obscure neural responses.

Purpose of the Study:

• To compare the Adjacent Response (ADJAR) algorithm with the General Linear Model (GLM) for EFRP estimation.
• To address the overlapping issue in EFRP estimation using a conjoint EEG and eye-tracking dataset.

Main Methods:

• Comparison of ADJAR and GLM algorithms on real EEG and eye-tracking data.
• Evaluation of different GLM configurations to account for overlapping potentials and differentiate fixation onsets.

Main Results:

• The ADJAR algorithm demonstrated overly restrictive assumptions for EFRP estimation.
• The GLM proved to be a more robust and efficient method for handling overlapping neural signals.

Conclusions:

• The General Linear Model (GLM) is superior to ADJAR for estimating Eye Fixation Related Potentials (EFRP).
• GLM configuration involves a trade-off between model assumptions and EFRP estimation quality, requiring control over estimated potentials to prevent high variance errors.