Clinical Validation of an Autonomous Online AI-based diagnostic system for the Detection of Diabetic Retinopathy

Diabetes is a burning problem that is on rise worldwide. By 2050 it is estimated that 1.31 billion people will be living with Diabetes Mellitus (DM), most of them with type 2 DM. In 2021, 95.4% of older adult diabetics had disability-adjusted life years (DALYs).  DR is one of the major complications of DM and is the leading cause of vision impairment both in developed and developing countries.  In a recent cross-sectional study conducted in 2021 that investigated the prevalence of DR in DM patients aged above 40 years, it was estimated that an alarming 21 million have vision impairment and 2.4 million are blind due to DR. Interestingly, undiagnosed DM had an odd’s ratio of 2 for vision-threatening DR. A systematic DR screening would reduce this burden to a larger extent.  Although high-income countries have successfully implemented DR screening programs, in countries like India it is partial.  Many nations use 2–4 fields fundus images, this has proven effective with 80–98% sensitivity and 86–100% specificity compared to the traditional seven-field evaluation or dilated clinical examination for DR.

While numerous DR algorithms exist, a majority have been validated using high-end, expensive fundus cameras. This reliance on costly equipment poses significant challenges, particularly in resource-limited settings and primary care centres, where access to such technology is often constrained. This limitation reduces patient participation in screening programs, hindering early detection efforts.

To address these barriers, the Oivi DR-AI system has been developed to work seamlessly with the Oivi fundus camera—a cost-effective alternative that can capture images quickly and efficiently.

Every algorithm developed must be validated within the intended population to demonstrate its efficacy and identify associated risks. Validation studies are crucial to establishing the clinical utility of the device in real-world settings. The findings from this study build trust, instil confidence among clinicians, and ultimately encourage the adoption of such innovations, ensuring their effective integration into everyday practice.