Research Project
9409553
Project Summary Over the past two decades, diabetic retinopathy (DR) has become the leading cause of adult blindness in the US, affecting 40% of all diabetic patients and resulting in $500 million a year in direct medical costs. Vision loss due to DR is largely preventable and can be reduced by up to 90% with appropriate eye screening. However, in the US, less than 50% of diabetic patients receive a recommended yearly eye exam due to many factors that include lack of access to eye care professionals. Distributed tele-ophthalmic screening thru primary care clinics can potentially provide all diabetic patients cost- effective, yearly evaluations to detect DR and prevent vision loss. However, gold-standard sensitive detection of DR using standard retinal photography is complex and cumbersome process requiring up to 7 images per eye. This screening process cannot for all practical purposes be achieved without having highly trained ophthalmic photographers. RetiVue proposes to develop the RetiVue DR in collaboration with Olympus, to create the first handheld, non- mydriatic, 160 field of view, widefield DR screening camera. It will allow single photo capture of an area up to ten times greater than conventional fundus cameras, allowing sensitive detection of DR at its earliest time points. Full integration of RetiVue and Olympus hardware will enable the most advanced and highest image quality handheld retina camera on the market. Use of automated alignment, auto laser focus, and auto image capture will allow complex imaging of the retina to be performed simply by positioning the iris, requiring no user knowledge of retinal anatomy. We have established clinical proof of concept with our patented technology, but require several additional innovations in optical design, automated image recognition, and retinal image processing to enable a commercial device. We will for this proposal optimize our alignment system and laser based focusing system for widefield imaging, allowing automated alignment and focus to image the retina before eye movement occurs. We will develop a new method of widefield, non-mydriatic peripheral retinal imaging using multiple LED slit-beam projectors to allow rapid, segmental, sequential image capture of 90°, 120°, and 160° FOV on diabetic patients. Finally, we will create the most advanced retinal image processing algorithms to remove Purkinje haze which prevents conventional cameras from imaging beyond 45° FOV and enable seamless stitching of segmental peripheral retina images into a single widefield image.
