I’m exploring whether it’s possible to detect real-time cognitive overload caused by prolonged reading or screen work — specifically when the brain–eye system is fatigued but the user keeps pushing.

The idea is to combine non-invasive signals: - eye movement patterns (saccades, fixation instability) - blink rate changes - pupil response - lightweight EEG bands (alpha / theta)

Not to diagnose anything, but to estimate a real-time cognitive load score that could say: “take a break now” before performance collapses.

Wearables today mostly analyze fatigue after the fact (sleep, HRV). This would try to detect overload as it happens.

Curious about: - prior work combining eye-tracking + EEG - known failure modes - feasibility outside a lab

Any pointers or “this won’t work because…” feedback welcome.