Video-based eye tracking for neuropsychiatric assessment

Area of Science:

• Neuroscience
• Biomedical Engineering
• Medical Technology

Background:

• Eye movements and pupillary changes are controlled by brain circuits.
• These eye metrics are sensitive to neurological and neuropsychiatric conditions.
• Video-based eye-tracking offers a non-invasive measurement approach.

Purpose of the Study:

• To present a video-based eye-tracking method using webcams for brain function assessment.
• To demonstrate the feasibility of using standard webcams for eye-tracking data acquisition.
• To explore the potential for early detection and remote monitoring of disorders.

Main Methods:

• Utilized mobile/laptop webcams for video-based eye-tracking.
• Collected data across 14 testing sessions, ~17,000 frames.
• Extracted computational and semantic features from eye-tracking videos.
• Employed a circular visual-tracking paradigm.

Main Results:

• Successfully extracted features from extensive video data.
• Demonstrated feasibility of webcam-based eye-tracking for psychomotor assessment.
• Identified specific eye-tracking features (radial/tangential variance) that predict psychomotor test performance.

Conclusions:

• Webcam-based eye-tracking is a viable tool for assessing brain function.
• This method holds potential for early diagnosis and serial monitoring of neurological conditions.
• Eye-tracking metrics can serve as biomarkers for psychomotor performance and related disorders.