Detection of Visual Fatigue Using Eye Tracking and a Wearable Device

Area of Science:

• Ophthalmology
• Human-Computer Interaction
• Biomedical Engineering

Background:

• Objective assessment of visual fatigue is crucial but challenging.
• Conventional eye-tracking methods have limitations in real-world applications.

Purpose of the Study:

• To explore objective markers for visual fatigue.
• To evaluate a novel spectacle frame with photodiodes for fatigue detection.
• To compare photodiode data with conventional eye-tracking.

Main Methods:

• Sixty participants performed a 30-minute reading task to induce visual fatigue.
• Subjective discomfort was measured using DESQ, OSDI, and SSQVF questionnaires.
• Eye movements were tracked using Eyelink, and periocular changes were measured by a photodiode-equipped spectacle frame.
• A time series transformer model classified fatigue states using both data sources.

Main Results:

• Subjective visual discomfort significantly increased post-task.
• Eye movement parameters like saccade amplitude/velocity and blink duration changed significantly; pupil size decreased.
• Fixation stability metrics correlated with subjective fatigue scores (SSQVF).
• The photodiode model achieved higher accuracy (82.5%) than eye-tracking (73.1%) in fatigue classification.

Conclusions:

• Photodiode-equipped wearable devices show significant potential for objective visual fatigue detection.
• This technology offers a practical alternative to laboratory-based eye-tracking for real-world monitoring.
• Fixation stability metrics are key objective markers correlating with subjective fatigue.