Medical artificial intelligence with a purpose

Artificial intelligence is quickly becoming a national obsession. With the Food and Drug Administration’s approval of AI-enabled devices for clinical use, AI is more real now than ever in medicine.

According to Carey Business School Professor Tinglong Dai, the real challenge of AI in medicine is figuring out how to embed AI devices and solutions into day-to-day health care workflows. At the same time, these revolutionary technologies must translate into real-world patient outcomes, improved productivity, and better access and equity to health care for vulnerable populations.

To meet these goals, Dai, Risa Wolf of the Johns Hopkins School of Medicine, and Haiyang Yang, Associate Professor of Marketing at Carey Business School, jointly investigate "Purposeful Design for AI-augmented Healthcare," a research effort supported by a $100,000 Johns Hopkins Discovery Award. They aim to develop broadly applicable principles and insights that will help seamlessly incorporate AI into the doctor’s office and other health care settings.

Johns Hopkins is a pioneer in the use of AI in clinical research and patient care, and Wolf and her team were among the first physicians to use LumineticsCore (previously known as IDx-DR), the first FDA-approved AI-enabled diagnostic device, in a pediatric care setting. Their early work demonstrated significant promise in using autonomous AI to reduce the cost of diabetic retinopathy screening, which is a complication of diabetes and a leading cause of blindness in working-age adults in the United States.

The annual screening rate for retinopathy is only about 15 percent in the U.S. One major hurdle is that patients do not find it convenient to schedule a separate visit and deal with a separate provider as well as the often tedious and soul-crushing billing process. An ongoing study at Johns Hopkins by Wolf, Dai, and colleagues found nearly 100 percent of patients completed recommended retinopathy screening using an AI option in their primary care office. The same study found just 22 percent completed retinopathy screening using traditional protocols.

“The true magic of using AI is its ability to allow health care to take place in settings other than a traditional setting,” said Dai. “When healthcare seems out of reach, AI emerges as a beacon, bridging the gap between patients and the healthcare they need.”

With the Discovery Award, Dai, Wolf, and Yang hope to build upon this success by scaling up in their collaborative investigation. Together, they launched the Johns Hopkins Workgroup on AI and Healthcare, which is part of the Hopkins Business of Health Initiative and aims to bring together all stakeholders in AI and health, including AI and medical innovators, researchers, providers, and policymakers.

Professor Yang, a behavioral scientist, added, “Understanding the psychology of the stakeholders is a prerequisite to effective utilization of medical artificial intelligence. Our research seeks to shed light on, for example, factors influencing health care providers’ adoption of AI systems. We also explore factors driving patients’ decision to receive AI-based medical care.”

The research is already halfway completed, and the trio anticipates that the grant will yield significant findings across multiple lines of integration, inspiring stakeholders in AI and health not only at Johns Hopkins but also on a national scale.