← Back to Explore

Computational Eye-Tracking Biomarker for Improved Neuropsychological Evaluation via Deep Learning

JSHS · 2020

Overview

Neil Milburn Plano West Senior High School Neuropsychological evaluation proves exceedingly valuable in applications that require a complex understanding of human behavioral interactions and mental processes, such as in the diagnosis of cognitive disorders. Current standards of neuropsychological assessments pose serious inaccuracies and difficulties in accessibility due to a degree of subjectivity present in monitoring behavior. This project uses computer vision and deep learning to tap into the oculomotor biomarker (eye movements) to create a non-invasive, low-cost, and effective approach to quantifying brain functions. The developed tool uses pupil-tracking and gaze-estimation algorithms to accurately assess the neurological and psychological deficiencies of patients. An infrared camera sensor (a) tracks eye movements and (b) estimates gaze points; in both cases, pattern recognition is employed to identify abnormal eye sequences in response to visual stimuli. Clinical testing on ADHD patients found that the tool was able to localize behavioral abnormalities within a 75-second test, including areas of inattention, poor reaction, and indecisiveness. The pattern recognition engine also classified neuromotor issues, such as a lagged movement of the eyes, rapid deviations of gaze, and eye tremors. With multiple unique patterns identified, the biomarker yields a 95% sensitivity rate and a 99% specificity rate for ADHD detection, proving this tool's capability in understanding human behavior, intent, and actions in cognitive disorders. The data suggests that the developed eye-tracking tool accurately classifies neuropsychological and motor issues, and the easily implementable nature increases viability in other applications that require an analysis of human interactions with their surroundings.

Awards (1)

  • 2nd Place Physical Sciences

Competition history

  • JSHS 2020 Category not listed

Resources

Related projects

Closest projects by meaning, across every fair and year in the corpus.

Source: Junior Science and Humanities Symposium

Save projects to your library

Sign in with Google to keep track of projects you find interesting, organized into folders. An account also raises your daily allowance for “Has this been done?”, and lets you create a key for the MCP server with a much higher limit than anonymous use. Browsing stays public.

Continue with Google