Transforming Glaucoma Detection

Dr. Ali Karimi and Dr. Aiyin Chen discuss data while looking a results on a computer.

AI Meets Electronic Health Records

New software that could revolutionize glaucoma screening is being developed at Oregon Health & Science University's Casey Eye Institute. Aiyin Chen, M.D., associate professor of ophthalmology, and Alireza (Ali) Karimi, Ph.D., assistant professor of ophthalmology, are working with a team to develop a program called GlaucoMate. The software, a machine learning-powered risk assessment tool, is designed to predict the probability that a patient will develop glaucoma based on data from their electronic health record (EHR).

Using machine learning to predict glaucoma risk

The GlaucoMate calculator takes standard EHR data—age, diabetes status, blood pressure, medication history, and other indicators of a patient’s health—and processes it through an AI model to screen for glaucoma risk. This allows basic screening without the need for an ophthalmologist to perform an eye exam or review imaging data.

The software promises to help overcome barriers to high-quality eye care for patients who lack access. Using GlaucoMate could significantly reduce the number of undiagnosed glaucoma cases, allowing for earlier intervention and potentially preventing vision loss for millions of people. When the tool is operational, providers will be able to input patient information and instantly view the likelihood of glaucoma risk, enabling targeted and efficient screenings by ophthalmologists.

The research group has developed the first iteration of the software using EHR data from patients at OHSU. It has achieved an area under the receiver operating characteristic (AUROC) curve of 0.67, indicating moderate predictive performance.

Addressing healthcare disparities

The Oregon Vision Health Network (OVHN), directed by Mitchell Brinks, M.D., M.P.H., is already improving access to high quality eye care across Oregon by creating a network of partners that provide screenings, telehealth services and training for local health workers to care for patients in their communities. The program now stations optical coherence tomography (OCT) machines, specialized equipment used for screening and diagnosing eye diseases, at federally qualified health centers (FQHCs) around Oregon. Images are then analyzed at the Casey Eye Institute. GlaucoMate can help the OVHN do even better.

"Screening efficiency is our challenge,” explains Dr. Chen. "In our high-risk populations, we currently screen 100 patients to identify 8 with glaucoma. This is a significant improvement over rates of glaucoma found in the general population—around 2.5%—but we can do better." Hence, the development of an AI tool that can tell health care providers which patients have the greatest need for screening.

A novel screening strategy

Traditional glaucoma screening requires specialized eye imaging and in-person ophthalmologist evaluations. These barriers prevent millions from accessing critical eye care. GlaucoMate uses the data entered for every patient at routine clinic visits, so even if someone has never seen an eye doctor, the software has enough information to help predict glaucoma risk. The implications for global health are enormous.

"The world's problem is that we can never have enough cameras and machines to screen everyone for glaucoma," Dr. Chen explains. "But patients already see primary care providers and have extensive medical records. We can plug that information into our calculator to narrow down who needs further evaluation."

An infographic depicting common health information collected in electronic health records, such as family history, medications, age, and blood pressure.

Early promise, significant impact

By pre-screening patients using existing electronic health record data, the system can enrich the screening population, making limited resources dramatically more effective.

But Dr. Chen's vision extends further. Her research also explores identifying "fast progressers"—patients whose glaucoma advances rapidly and who could benefit from early intervention. By understanding systemic health factors contributing to eye disease progression, her team aims to develop targeted treatments beyond traditional intraocular pressure (IOP) management.

A team approach to science

Dr. Chen is quick to credit an interdisciplinary team for developing this innovation. “This is the era of team science,” she says. “I work with Dr. Karimi, a bioengineer who developed the GlaucoMate software, and Wei-Chun Lin, M.D., Ph.D., a senior data scientist who specializes in AI and building models. The OVHN director, Dr. Mitch Brinks, oversees implementation of the programs and screening models we use now.”

Health center staff gather data, and IT specialists build and maintain the infrastructure needed to transfer ocular images from community health centers to Casey Eye Institute. Dr. Chen belongs to the Glaucoma Informatics Group at Casey Eye Institute, where “we share ideas and data to collaborate,“ concluding, “It takes a whole team to do something like this.”

From eye health to whole health

Finally, Dr. Chen emphasizes the importance of a holistic perspective on health. “I’m an ophthalmologist, but I don’t care only about my patients’ eyes,” she says. “As eye care providers, we need to recognize that eye diseases are deeply connected to overall systemic health.”

Recent data indicates that eye diseases such as glaucoma and macular degeneration are linked to overall health. “We tend to zoom in on the eye,” says Dr. Chen. “But reviewing a patient’s full history can shed light on why their eye condition is changing, for example.”

Although continued development and refinement of GlaucoMate is ongoing, this technology represents an innovative approach to eye screening. Leveraging patient data promises to make glaucoma screening more efficient and accessible, particularly for those in rural or underserved areas who face barriers to specialist care. The project demonstrates the power of interdisciplinary collaboration at Casey Eye Institute and OHSU. It also reinforcing the connection between eye health and overall wellbeing and providing a model for future health care innovation and research.