Pivotal Trial of an Automated AI-based System for Early Diagnosis and Prediction of Late Age-related Macular Degeneration

The purpose of this study is to perform a pivotal trial of iPredict, an automated AI-based system for early diagnosis and prediction of late AMD in primary care and ophthalmology settings. Patients will be invited to participate in this study by having non-dilated photos of their eyes taken by an FDA approved fundus camera (DRSPlus from Centervue Inc., CA), at their primary care doctor's office or general ophthalmologist office. The photos will then be transmitted securely and analyzed by computer in the cloud (telemedicine features). Sufficient accuracy of the automatic system has been established compared to the ophthalmologist's diagnosis. In this study, we aim to validate the system against the prospectively taken OCT image and color fundus images.

Background AMD affects 15 million Americans, with 200,000 new cases diagnosed each year. At present, there is no treatment for dry AMD. Besides blindness, AMD has other indirect complications such as depression, social dependency, and the risk of fall and injury. The prevalence of this disease is expected to grow substantially as life expectancy continues to increase and record numbers of Baby Boomers enter their senior years. The total direct cost of AMD is $220 billion per year and is expected to increase ~1.5 fold. The Age-Related Eye Disease Study (AREDS) showed that specific vitamin supplementation protocols can reduce the risk of progression from intermediate to late AMD by ~25% which in turn could lower the cost of AMD 17.6% if fully implemented. To accomplish this, it is crucial to perform large scale population screening to identify the individuals with early- or intermediate-stage of AMD and better predict those at risk of developing late AMD, but such a system is currently not available. Although articles have been published on automatic AMD pathology detection, none of these systems are available for screening due to lack of validation and commercial readiness. Considering this urgent need, we aim to develop an automated tool iPredict for early diagnosis and prediction of AMD, and make it widely available in both urban and remote/rural areas and for large- scale screening (through its telemedicine features), and thereby prevent blindness.

Primary and Secondary Study Endpoints The accuracy of the iPredict software developed by iHealthScreen system in early diagnosis of AMD using color retinal photos vs. that of human expert graders for AMD. Also, the prediction of late AMD progression in 1 or two years.