Approved Research

Discovery and development of retinal imaging-based diagnostic, prognostic and predictive biomarkers of ophthalmic and non-ophthalmic diseases with retinal manifestation

Lay summary

The study aims to detect changes inside the eye (retina) in various eye diseases and potentially other diseases with changes in the retina utilizing advanced and intelligent computer programming tools known as artificial intelligence or AI. These AI-enabled tools are expected to identify and quantify changes observed in the retina photographs automatically, and the readouts obtained can be highly reproducible. These readouts, also known as biomarkers, can be reliably used by a physician for identifying and managing eye diseases. Additionally, these biomarkers can potentially help to select patients and quantify treatment responses to support the investigation of new pharmaceutical drugs. We also envision these AI-enabled tools can provide new biomarkers by combining information from multiple types of retina photographs and other disease-related information.

We expect that this project will last about three years. Over these three years, we hope to develop a range of AI-enabled tools for diseases that are leading causes of vision loss, such as age-related macular degeneration (AMD) and diabetic macular edema (DME). In addition to AMD and DME, we also recognize a huge potential for these tools in other diseases, such as Alzheimer's disease, where changes in the retina can happen at the early stage. Since retina photographs are relatively simple to take and do not involve invasive procedures, these AI-enabled tools are expected to offer immense value to public health.