News

Artificial Intelligence (AI) and Machine Learning are transforming healthcare by helping physicians and other clinicians manage growing demands at the point of care.

“AI is embedded in electronic health record systems, provides computerized decision support, and offers alerts and reminders to clinicians,” said Omolola Ogunyemi, PhD, professor and department chair of Preventive and Social Medicine at Charles R. Drew University of Medicine and Science (CDU). “It’s a tool that can help us get better at certain things, but it is not intended to be a replacement for human beings.”

According to Ogunyemi, advances in AI and machine learning have shown particular promise in diagnosing health conditions, such as diabetic retinopathy.

Diabetic retinopathy, a complication of diabetes that damages the blood vessels of the retina, remains the leading cause of blindness among working-age adults in the United States. Although the condition is largely preventable with early detection and timely treatment, it continues to disproportionately impact individuals in medically underserved communities, where access to specialty care is often limited.

“In underserved settings, AI can be useful in addressing conditions that are prevalent in those communities but often go under-managed because of limited resources and high patient volume,” said Ogunyemi. “By making providers more efficient and reducing time spent on documentation, AI can free up more time to focus on patient care.” 

Ogunyemi recalls one such instance from her early years as a researcher at CDU, when she witnessed firsthand the consequences of delayed specialty referrals in safety-net settings.

“A patient at a federally qualified health center had diabetes and was not managing it well,” recalled Ogunyemi. “His primary care provider referred him to the County for specialty care, but the clinics were inundated. After waiting for months to be seen, he went blind.”

Motivated by this experience, Ogunyemi helped secure research funding from the National Institutes of Health (NIH) to explore tele-retinal screenings, an approach that placed digital non-mydriatic cameras in federally qualified health centers to capture high-resolution images of the retina within primary care settings.

“AI driven models provide another way to manage care when resources are scarce,” explained Ogunyemi. “Clinicians are able to be more intentional about who needs care urgently.”

She also emphasized that ensuring AI tools work effectively and equitably requires careful attention to how models are developed and trained.

“If you take a model trained on data from one health system and apply it to a completely different community, you risk underdiagnosing or missing conditions altogether,” she said. “One strategy is to develop models using data that broadly reflects the US population. Another strategy, which may be important in the healthcare safety net, is building AI that’s targeted for specific populations, using data that reflects their lived realities and health burdens.”

Universities like CDU play a critical role in shaping the ethical and equitable use of AI in medicine. In addition to training future health professionals through programs such as the forthcoming Master of Science program in Health Informatics, CDU helps ensure that underserved communities are represented in research, without exploitation.

“Our mission calls us to advocate for under-resourced communities while doing original research that leads to meaningful change,” Ogunyemi said. “That means making sure AI is used to close gaps in care, not widen them.”