Do pupil-based binocular video eye trackers reliably measure vergence?

Area of Science:

• Ophthalmology
• Visual Neuroscience
• Biomedical Engineering

Background:

• Accurate binocular eye tracking is crucial for determining vergence, fixation disparity, and 3D fixation distances.
• These metrics are vital for research in visual perception, reading dynamics, and attention.
• Previous studies suggest pupil-size variations can introduce significant errors in eye-tracking data.

Purpose of the Study:

• To evaluate the accuracy of a high-end binocular video eye tracker (SR Research EyeLink 1000 plus) in measuring binocular parameters.
• To investigate if pupil-size fluctuations cause apparent vergence errors independent of actual eye rotation.
• To determine if pupil-based eye trackers are suitable for precise binocular measurements.

Main Methods:

• Participants fixated a target at 77 cm for 160 seconds.
• Light intensity was varied to induce pupil-size changes.
• Eye movements were monitored using a binocular video eye tracker, analyzing pupil and corneal reflection (CR) signals.

Main Results:

• A decrease in reported horizontal vergence (up to 2 degrees) was observed with increasing pupil size in most participants.
• Corneal reflection (CR) signals showed no systematic movement, indicating no actual eyeball rotation.
• The observed vergence changes were attributed to pupil-size artifacts, not physiological eye movements.

Conclusions:

• Binocular video eye trackers relying on pupil-CR or dark pupil techniques are not sufficiently accurate for measuring vergence.
• These trackers may produce unreliable data for determining fixation disparity and 3D fixation distances.
• The findings highlight limitations of current pupil-based eye trackers in demanding binocular vision research.