Issue link: https://www.e-digitaleditions.com/i/1501973
F thoughts after each unit of study. One found the unit on the safety of medications to be particularly overwhelming. "I caught a glimpse of how far down the rabbit hole actually goes," the student reported. "There are so many layers of information to digest—from drug/device approval to code-based algorithms to neural- network processing engines. Each of these ideas could easily have an entire degree based around them. My biggest takeaway from this unit was the sheer magnitude of data, analysis and cooperation that is required for AI to be a viable technology." Chang emphasizes that students must still be schooled in evidence-based medicine. But they can now incorporate this new AI skill set into their decision making to question whether a program is producing reasonable results. Sreedhar, who graduated from a medical school in India that had only one computer in its library, stresses that students need not have more than a basic knowledge of math and science to take the course. "We are not training them to be computer science engineers," she says. From the beginning of her career, Sreedhar has worked to improve the ability of students and residents to access the literature that guides treatment. "Students have so much to learn, and time is so limited," she says. "At some point, it became very clear that there was no way that doctors could keep up with the medical literature. AI and ML can do some of that research for them." 28 | Summer 2023 Grounded in Reality: Artificial Intelligence in Medical Education But there may be no other way forward, according to Linda Chang, PharmD '97, MPH '05, clinical associate professor of family medicine at the Rockford campus. "We're trying to train physicians for far into the 21st century," says Chang. With that goal in mind, Chang teamed up with Radhika Sreedhar, MD, MS, FACP, associate professor of clinical medicine in Chicago and director of curricular integration, to develop an elective AI course for fourth-year medical students. Created with the help of colleagues with expertise in AI, informatics and instructional design, the Artificial (Augmented) Intelligence in Medicine course introduces students to AI by examining biomedical informatics foundations, the essentials of healthcare data science, the use of predictive analytics and artificial intelligence, and ways to integrate Al into their future practices. "Students look at what's behind each of the AI programs and learn how they can use technology to provide patient care in an equitable way," Chang explains. "The course covers all kinds of common patient- case scenarios. And, once students learn the background of a case, they can practice their AI skills through weekly real-life scenarios. "Among other things, Dr. Sreedhar and I help them spot what could potentially be a problem if data isn't used correctly and recognize when there might be health-disparity gaps between different populations," says Chang. After the course was introduced in fall 2019, students anonymously shared their FROM THE FIRST DAY OF CLASS, medical students are inundated with information. There's so much to learn and so little time. How can they possibly absorb it all—and learn how to use artificial intelligence in their future medical practice as well? Linda Chang, PharmD, MPH