Publication

Abstract : It used to be quite challenging to diagnose cognitive impairment simply by visiting a geriatrician. Diagnosing cognitive impairments, particularly in the age group of 18 to 22, has become increasingly important as the prevalence of these conditions continues to rise. Traditional diagnostic methods were often time-consuming and expensive, hindering timely intervention. To address this, we propose a model EYE - COG: utilizing eye-tracking technology to detect cognitive impairments by monitoring real-time eye movement during a Trail Making Test. Our model leverages various machine learning and deep learning techniques to analyze features fixation, saccade, and blink. By employing CNN model, different stages of cognitive impairment can be accurately classified. Model EYE- COG, utilizing early stopping with optimization, achieves an impressive accuracy of 83%