Screening for Diabetic Retinopathy by Image Content

	An example of a red-free fundus image that has been automatically analyzed to show important structures such as the vasculature, location of the optic disc, the macula region, and lesions.

The World Health Organization estimates that 135 million people have diabetes mellitus worldwide. Consequently, diabetic retinopathy (DR) is the leading cause of new blindness in working-age adults in the industrialized world [1, 2]. It is estimated that as much as $167 million dollars and 71,000-85,000 sight-years could be saved annually in the U.S. alone with improved retinal screening methods [3].

The current model for ocular telehealth is the reading center, e.g.,
Joslin Vision Network, Boston, MA; Fundus Photograph Reading Center, University of Wisconsin, Madison, WI; and the Vanderbilt Ophthalmic Imaging Center, Nashville, TN. Reading centers provide a mechanism for collecting and managing fundus photography for clinical studies and they are beginning to support broad-based screening activities (e.g., through the Veterans Administration). The current reading center model uses human readers to screen patients, therefore limiting their ability to provide high-throughout, rapid screening of large populations. Typical turn-around time for a reading center today is 24-72 hrs. To improve throughput, reading centers are beginning to investigate the use of computer-assisted screening technologies.

Through this research, we are developing a Retinopathy Image Search and Analysis (RISA) technology that uses a content-based image retrieval (CBIR) method to perform rapid analysis and diagnosis of digital retinal imagery through a telemedicine model. Our goal is to facilitate the screening of much larger populations of people that are at risk today of diabetic retinopathy by providing a high-throughput, low-cost computer-based diagnostics method that can be performed by non-experts using inexpensive non-mydriatic fundus cameras.

Acknowledgements

This research is being supported by the National Eye Institute R01 EY017065 of the National Institutes of Health, the Plough Foundation, Memphis, Tennessee, and Research to Prevent Blindness, New York, New York, along with the Laboratory Directed Research and Development Program of the Oak Ridge National Laboratory.

	Research to Prevent Blindness

References

[1] J.C. Javitt, L.P. Aiello, L.J. Bassi , Chiang YP, Canner JK., "Detecting and treating retinopathy in patients with type I diabetes mellitus: Savings associated with improved implementation of current guidelines," American Academy of Ophthalmology. Ophthalmology. 1991 Oct;98(10):1565-73.
[2] Centers for Disease Control and Prevention (2003) National Diabetes Fact Sheet. (http://www.cdc.gov).
[3] M. Larsen, J. Godt, M. Grunkin, “Automated detection of diabetic retinopathy in a fundus photographic screening population,” Invest Ophthal Vis Sci 2003, 44:767-71.