Illinois Medicine

of Biomedical Engineering and the Department of Ophthalmology and Visual Sciences. "We currently have about 350 patients with sickle cell disease who are seen in ophthalmology, which, left undetected and treated, can sometimes result in visual loss," says Lim, the Marion H. Schenk, Esq., Chair in Ophthalmology for Research of the Aging Eye; vice chair for diversity, equity and inclusion; and director of the Retina Service. "So we are using AI technology to look at the ocular blood flow of sickle cell patients. While this is still in the research phase, we've found that AI data correlated with the stage of sickle cell disease. This can be helpful in creating AI systems for staging sickle cell retinopathy." Lim also points out a study done at UIC and other centers using a now FDA-approved AI system for screening diabetic patients for referable diabetic retinopathy. Lim says this study was significant because it tested the utility of the algorithm on UI Health's diverse patient population—and it was found to be able to detect referable retinopathy. Thus, the final product is applicable to UI Health's population. "An AI system is only as good as the data that went into its development," says Lim. "Data from diverse populations is required during development and testing for the AI system to be applicable to diverse populations. "Hopefully, this will help decrease the disparities of care for inner-city patients, many of whom have difficulty accessing care," Lim adds. "With diabetes being the No. 1 cause of blindness in the working-age population, a simple and effective solution could be AI-directed point-of-care eye screening in a primary care clinic. If you can use AI screening for at-risk patients, you can identify those needing further testing and treatment." Because AI employs objective measures to create more consistent diagnoses, the technology can be useful as a training tool for newer physicians, especially when there's disagreement about a diagnosis. "AI can provide another data point for physicians to use along with all of the other diagnostic tools at their disposal," explains Chan. "All of the available technology only provides information; a clinician's judgment is still critically important for the care of patients." Chan notes that UIC is well positioned to ensure that it is providing access to care, especially in resource-limited settings. Accordingly, AI's use has the ability to drive down costs by providing more cost-effective and efficient healthcare. "Many patients, for instance, experience some very real challenges just getting to a clinic for in-person visits," says Chan. "If systems with better predictive models are created, these systems may actually decrease the number of times that some patients may need to visit a clinic in person, which would help alleviate patient stress. Using AI technology as a diagnostic and educational tool will advance our understanding of this technology and benefit all of our patients." As Chan and his colleagues continue to explore the potential of artificial intelligence in ophthalmology, their research and educational endeavors are getting a welcome boost from the family of former patient Gerhard Cless. The Cless Family Foundation has generously funded vision research, patient care and education at the College of Medicine—including a recent gift to advance the work of the Ai-O in the Department of Ophthalmology and Visual Sciences. 24 | Summer 2023 Jennifer I. Lim, MD, FARVO " " An AI system is only as good as the data that went into its development. Data from diverse populations is required during development and testing for the AI system to be applicable to diverse populations. P H O T O : D I A N E S M U T N Y

• Cover