The potential of AI in retina with Daniela Ferrara

This conversation features Daniella Ferrara, Chief Medical Officer for TopCon Healthcare, discussing the transformative potential of artificial intelligence in retinal imaging and ophthalmology. Ferrara, an ophthalmologist and retina specialist with 25 years of experience in retinal imaging innovation.

The discussion centers around a panel Ferrara moderated at the OIS Retina event, which precedes the ASRS meeting. This panel brought together experts from pharmaceutical companies, AI algorithm developers, and clinicians to explore how artificial intelligence can revolutionize clinical trials in retinal medicine.

Ferrara identifies 2 critical problems that AI aims to solve across different healthcare ecosystems. In drug development, AI can accelerate bringing new therapies to patients by increasing success rates, reducing timelines, and improving resource efficiency. In clinical practice, AI enables more informed decision-making and supports personalized treatment approaches. She describes 2 main categories of AI algorithms currently being developed. Segmentation algorithms automate tasks traditionally performed by human experts, such as identifying and measuring biomarkers in retinal images over time, providing efficiency and cost benefits. Future prediction algorithms are more complex, analyzing retinal images to predict the future state of the retina, sometimes using identifiable biomarkers and other times operating as "black box" systems.

The conversation highlights concrete implementations of AI in clinical trials, including accelerated patient screening, improved recruitment processes, and enhanced interpretation of trial results. Some emerging endpoints now require AI for proper image analysis, marking a shift from theoretical applications to practical implementation.

A particularly exciting development discussed is "ocolomics" research, which uses retinal images to predict systemic diseases. While clinicians have traditionally diagnosed conditions like diabetes and hypertension through retinal examination, AI dramatically enhances this capability. Preliminary research shows promise for predicting neurodegenerative diseases such as Alzheimer disease, MS, and Parkinson disease, as well as improving predictions for cardiovascular and metabolic conditions.

Ferrara emphasizes that the field has moved beyond "pie in the sky" concepts to concrete, practical applications. The integration of AI in retinal imaging represents a significant paradigm shift, offering the potential to transform both clinical trial methodology and patient care delivery. The technology's ability to meet patients where they are and provide more accurate, efficient diagnostics positions AI as a crucial tool for the future of ophthalmology and broader healthcare applications.