Pediatric Perimeter is a project undertaken at the Engineering the Eye Workshop organized by Srujana Innovation Center, LVPEI and MIT Media Lab. The project's aim is to automatically acknowledge a child's reaction to a visual stimuli provided in the Pediatric Perimeter device. This repo contains a Segmentation algorithm deployed in Pediatric Perimeter, being developed by Srujana Innovation Center, MIT Media Labs and LVPEI Hospital, to track gaze (pupil tracking) of babies in real-time using Viola-Jones Face Detection algorithm and Fast Radial Symmetry Transform.
Problem Statement: Pediatric Perimeter is a novel, first of its kind device to measure the side vision of infants. The procedure of quantifying visual field is called perimetry. This device is used on children with potential disease to cause a visual field defect such as glaucoma, retinitis pigmentosa, Retinopathy of Prematurity,etc. It involves checking the eye/head movement of an infant to a visual stimulus in particular areas in the visual field using various tests in perimetry involving LEDs fitted inside the Pediatric Perimeter. Currently, the clinician has to acknowledge the reaction from the children during testing inside a Pediatric perimeter. Doing it manually leads to error in the reaction time calculation. Automating this process will help in reducing this human error and help in getting more accurate readings.
Synopsis: “From the video of baby's response recorded using Pediatric Perimeter prototype, this algorithm checks whether and when the subject has reacted to the given stimulus in real time.”