PITA Fiscal Year 2008 Projects - Biomedical and Health Engineering

Detection and Screening for Diabetic Retinopathy by Real Time Pupil Size Monitoring

Principal Investigators: Ozan Tonguz, Jelena Kovacevic

Approximately 4.1 million US adults 40 years and older have diabetic retinopathy (DR); 1 out of every 12 persons with diabetes mellitus (DM) in this age group has advanced, vision-threatening retinopathy, which can both lead to loss of independence and reduced quality of life.

Automated image analysis tools have the potential to play an important role in the diagnosis and treatment of DR. The primary effect of DR is that it can cause vascular damage on the retina before the patient develops more severe symptoms. The damage on the retina is not uniformly distributed, that is, the relation of pupil response to light stimulus on the center of the retina and on the periphery of the retina will be different for a patient with DR and a healthy person. This difference can be exploited to design a system to be used for early detection of DR. This approach of using pupil response as a biosensor for retinal function has been studied and validated.

We have been working with Neuro Kinetics, Inc., a Pittsburgh-based company, and have developed an automated, high-speed eye tracking system with robust pupil center estimation that has the potential to be used as a general platform for early diagnosis of DR.

We propose to develop a digital image based pupillometry and analyzing system that utilizes our already functioning eye tracking system to screen for diabetic retinopathy.

Our ultimate goal is for the system we develop to be used to perform tests on the general population. Therefore, our system will facilitate a rapid test, be robust to interference, and provide accurate information. Compared to existing methods, our methodology will provide an objective test result, will be non-invasive, easy to use, and will not require extensive training for operators.