Computer Vision and Eyetracking Approach for Assessing Visual Disorders in Premature Infants

Abstract

Premature infants are subjected to an elevated risk of abnormal visual development, leading to potential visual impairments such as field defects, optic nerve dysfunctions, eye movement abnormalities, and nystagmus. To mitigate these dangers, early and exact assessment is important. This paper presents a computer vision approach that uses face mapping and eye-tracking technologies to objectively and accurately assess further visual disorders in infants. Our methodology integrates eye-tracking with facial recognition techniques to quantify eye movements and gaze stability accurately. Detailed video recordings of infants responses to visual stimuli are analyzed using a custom-developed algorithm, which processes eye and head movements to generate precise metrics of visual function. The results demonstrate that our computational tool significantly enhances the diagnostic capabilities by providing quantifiable and objective metrics that supports the clinical observations, thus assisting healthcare professionals in identifying and addressing potential visual abnormalities at early stages effectively.