Dr. Foo Li Lian and Dr. Kristie Nguyen share how this revolutionary tool is poised to redefine the industry.
What if a simple eye scan could reveal more than just your eye health? This is the promise of RETFound. By optimizing the power of deep learning and vast datasets of retinal images, RETFound is ushering in a new era of predictive and personalized eye care.
RETFound, an advanced artificial intelligence (AI) model developed by researchers at Moorfields Eye Hospital (UK) and the UCL Institute of Ophthalmology (UK), is empowering clinicians to see more, understand better and care smarter.
This breakthrough tool leverages retinal imaging to detect not only eye diseases but also systemic health risks such as glaucoma, Parkinson’s disease and heart attacks. RETFound represents a significant leap in predictive diagnostics, enabling earlier intervention and personalized care.
A 2024 study* showed that the RETFound-enhanced deep learning model significantly improves accuracy and net benefit in community eye disease screening, offering a promising solution for global eye health challenges, especially in low and middle-income countries.
Decoding the retina’s secrets
RETFound is a cutting-edge AI model designed to analyze retinal scans and extract vital health information. Retinal imaging offers a non-invasive view of a patient’s vascular and neurological health. By detecting subtle patterns and markers in the retina, RETFound can predict the likelihood of developing severe systemic conditions.
As one of the first AI foundation models in healthcare—and the very first in ophthalmology—it was developed using millions of eye scans from the UK’s National Health Service (NHS).
“RETFound demonstrates significant potential in the early detection of glaucoma and other ocular diseases by identifying subtle retinal features that may not be readily apparent, even to trained professionals during fundus examinations,” said Dr. Foo Li Lian, a consultant with the Cataract and Comprehensive Ophthalmology Department at the Singapore National Eye Centre. “This capability could enable earlier interventions, which are crucial for conditions like glaucoma, where vision loss is irreversible and, if left undiagnosed, could eventually lead to blindness.”
The retina often reflects systemic health conditions, including diabetes, hypertension, and neurodegenerative diseases. “AI models like RETFound enhance early detection by identifying diagnostic markers on retinal images. This advancement positions eye care professionals as key contributors to systemic disease diagnosis and screening, facilitating timely interventions and improving overall health outcomes,” she added.
“With the sheer amount of data RETFound analyzes to understand ocular pathology and its connection to systemic health, diagnostics can only benefit from this technology,” emphasized Dr. Kristie Nguyen, a board-certified optometrist from Florida, USA. “By comparing numerous images in its database, it will greatly assist clinicians in detecting early changes in systemic diseases during routine eye exams.”
Clinical implications in optometry
The integration of AI technologies like RETFound is poised to redefine the role of optometry, transitioning it from reactive care to a more predictive, preventative and personalized approach.
“These technologies could broaden the scope of optometry, enabling practitioners to address both ocular and systemic health issues, thus positioning the profession as a vital component of holistic healthcare delivery,” said Dr. Foo.
A significant advantage of AI tools lies in their ability to streamline patient management. “AI tools hold great promise in reducing the referral burden by enabling optometrists to triage cases more effectively, ensuring that only patients requiring specialist attention are referred,” Dr. Foo explained. “However, the accuracy and sensitivity of these tools must be robust to prevent over reliance or under-referral.”
Dr. Foo further emphasized the importance of equipping optometrists to maximize the potential of AI tools like RETFound.
“Optometrists will play a critical role in adopting RETFound, converting AI-generated insights into impactful patient care. As AI technologies like RETFound become increasingly integrated into clinical practice, optometrists will need to develop a thorough understanding of the technology, allowing them to confidently interpret AI findings and assess their clinical relevance,” she explained. “Beyond interpreting AI outputs, optometrists will also be tasked with effectively communicating these results to patients and ensuring timely referrals to other healthcare providers when necessary.”
Revolutionizing patient care and engagement
Highlighting the impact of RETFound in patient care, Dr. Nguyen remarked: “I think it will be a step in the right direction to offer optometrists more options in better patient care and management. Allowing patients to compare themselves to the population with a suggested prognosis/diagnosis could boost morale when it comes to treatment.”
Echoing this perspective, Dr. Foo elaborated on how RETFound enhances patient engagement.
“RETFound has the potential to revolutionize patient care by providing clear, data-driven insights into a person’s health, helping them better understand their condition,” she said. “However, building patient trust requires transparency—patients need to know how AI contributes to diagnoses and care decisions. It’s essential to address concerns about accuracy, as well as the risks of false positives or negatives, which could lead to over-referrals or missed diagnoses. Clear communication and patient education are crucial to fostering confidence in AI-driven tools.”
Hurdles and learning curves
Dr. Foo also acknowledged the challenges of integrating RETFound into optometry practices.
“Incorporating RETFound presents several hurdles,” she noted. “First, significant investment in advanced imaging equipment will be necessary to generate high-quality retinal images suitable for AI analysis. Additionally, comprehensive upskilling and the establishment of standardized protocols will be essential for optometrists to accurately interpret AI outputs and ensure timely referrals to appropriate healthcare providers. Seamless integration into existing workflows will also be needed to minimize disruptions to practice operations. Addressing these challenges is crucial to fully realizing RETFound’s potential in enhancing patient care,” she explained.
Dr. Nguyen added, “There may be learning curves in using the software. But with time, practice will make it easier to adapt. Continuing to expand the database is also important to capture a wider variety of presentations and scenarios, demonstrating how systemic health can affect the eyes and predicting future outcomes,” she said.
A new era of predictive healthcare
Despite these challenges, RETFound represents a significant advancement in optometry, transforming how eye care professionals diagnose and manage both ocular and systemic health holistically.
With continued innovation, education and integration, RETFound promises not only to elevate the role of optometry but also to lead us into a new era of predictive and personalized medicine.
Editor’s Note: A version of this article was first published in COOKIE magazine Issue 18.
Reference
- Zhang J, Lin S, Cheng T, et al. RETFound-enhanced community-based fundus disease screening: Real-world evidence and decision curve analysis. NPJ Digit Med. 2024;7(1):108.
