A Novel Image Based Screening System for Early Detection of Age-related Macular Degeneration

Period of Performance: 09/01/2016 - 08/31/2017

$241K

Phase 1 SBIR

Recipient Firm

Ihealthscreen, Inc.
CHARLOTTESVILLE, VA 22901
Principal Investigator

Abstract

PROJECT SUMMARY In this SBIR Phase I proposal, we aim to develop a retinal imaging based automated screening system to identify individuals at an early or intermediate stage of age-related macular degeneration (AMD) and thus enable treatment to prevent late AMD. Late AMD, in the dry or wet form, is the leading cause of vision loss in the developed countries. The Age-Related Eye Disease Study (AREDS) showed that specific antioxidant vitamin supplementation reduces the risk of progression from intermediate stages to late AMD and maintains visual acuity in approximately 25% of patients. Indeed, there is no present treatment for dry advanced AMD. While treatment of wet AMD with Intraocular injections can be effective in maintaining vision, such treatments are costly, and may be associated with significant cardiovascular risks, or even progression of dry AMD. While the direct impact of AMD is vision loss, it has other indirect complications such as depression, social dependency and risk of fall and injury. Hence, it is critical to identify patients at the earlier stages. Unfortunately, there is no effective, automated large scale screening tool to accomplish this, and the patients themselves may be asymptomatic. The goal of this proposal is to provide such a tool. The current few screening systems are mainly semi-automatic and rely on color fundus (CF) photography to identify drusen, the classic lesions of early AMD. However, no screening system captures Reticular Pseudo Drusen (RPD), a recently found and common early AMD lesion that carries an even worse prognosis for late AMD than ordinary drusen. Therefore, we will utilize both CF photography for drusen and high contrast Red-free (RF) imaging for RPD detection in a multimodal automatic screening tool to detect both forms of early AMD with high accuracy. Furthermore, lack of ophthalmologists in underserved areas may preclude the timely detection of early AMD. Therefore, we will next incorporate our screening tool in a telemedicine platform that can be deployed in such underserved areas to identify patients with AMD for referral to an ophthalmologist. In the current proposal, we will develop and validate the necessary algorithms to demonstrate that an automated screening system for early/intermediate AMD is feasible. In our Phase II SBIR proposal we will develop and test the complete system along with CF and RF retinal image capture devices that can be suitable for use of our system in a telemedicine platform.