Eye-Tracking Technologies for Cognitive Assessment After Acquired Brain Injury: Systematic Review.

The review included 27 studies with 872 participants, but 15 of 27 studies had fewer than 50 participants and only 2 had 100 or more. Most studies were cross-sectional and underpowered, and no quantitative pooling was possible. The evidence was dominated by mild traumatic brain injury and concussion, so it says little about stroke, hypoxic-ischemic injury, severe traumatic brain injury, or disorders of consciousness. Paradigms, devices, metrics, and reference standards varied widely, so results cannot be applied to any single eye-tracking test. Only 1 of 27 studies explicitly reported no harms, and 13 of 27 reported some feasibility or exclusion information, so safety and practicality are not established. Classification studies often used the same data to select predictors and thresholds and lacked external validation, so reported accuracy may be overestimated. Pediatric evidence was limited to one study of mild closed head injury.

Funded by Current Research Funds 2026, Ministry of Health, Italy. Authors declared no conflicts of interest.

Do not use eye tracking as a stand-alone cognitive assessment after acquired brain injury. The evidence was very low certainty for most domains, and the review did not show routine superiority over conventional neuropsychological testing.