Cornelius James, MD, is a Clinical Assistant Professor of Internal Medicine and Pediatrics. He leads the Data Augmented Technology Assisted Medical Decision-making (DATA-MD) curriculum initiative, and is an Educational Advisor to Precision Health’s Education & Training Workgroup.

What are your research interests?
I am interested in medical education research. More specifically, I am interested in teaching learners across the medical education continuum clinical reasoning and evidence-based medicine (EBM). Recognizing the impact that AI/ML and precision health will play in clinical reasoning, I have begun to explore how to incorporate this content into medical education curricula.

Please talk a little about your work with Precision Health’s Education & Training and Implementation workgroups. What have your contributions been?
It has been a pleasure working with the Education & Training and Implementation workgroups. During the 2021-22 academic year, we have developed a series of webinars to begin to educate frontline clinicians about the role of precision health and AI/ML in healthcare. This series has been very successful as it has allowed learners, faculty, and staff (locally and from around the country) to hear talks from experts at U-M about topics that will change the way health care is delivered, and more than likely the way we are teaching future health care providers and other health care roles. This will culminate with an in-person symposium on March 16, 2022, where we will have experts at U-M deliver talks and participate in panels, and we will have a leader and pioneer in the field of digital medicine [Robert M. Wachter, MD] deliver the keynote address.

Can you talk about the DATA-MD curriculum, which incorporates ML implementation and physician training, and what you hope to accomplish with it?
Frontline clinicians will be key stakeholders as end-users of AI/ML models. The DATA-MD curriculum is certainly educational programming, but I also view it as a movement toward the health care of the future. It is designed to teach clinicians to use AI/ML to augment medical decision making. EBM is a proven framework for evaluating and applying results from research. Given the overlap between EBM and AI/ML, we are using EBM as a framework to teach learners about the strengths and limitations of AI/ML in health care, and how to apply model outputs. For example, most of our current work is focused on using AI/ML to improve the diagnostic process and reduce diagnostic errors.

I have long been concerned about the “hidden curriculum” in health profession education. To address this, we are developing educational programming that targets learners across the medical education continuum. This approach will be conducive to allowing learners and teachers to begin to speak the same language and to some extent start at the same place. As AI/ML become more ubiquitous in health care, I expect that we will begin to provide educational programming that is more targeted to a specific learner level. Ultimately, we are looking to provide the tools necessary for clinicians to be engaged stakeholders when it comes to development and deployment of AI/ML-based technologies in health care.

The DATA-MD team comprises diverse backgrounds and areas of study. Why is such an interdisciplinary approach so important?
I am working with an amazing group of faculty, students, residents, and staff. Many schools and departments are represented on our team. It includes clinicians, researchers, lawyers, computer scientists, pharmacists, medical educators, nurses, and librarians. Because of the impact that AI/ML will have on health care, an interdisciplinary approach is essential. There are so many layers to developing, implementing, and teaching this content. Just as development of health care AI/ML models requires an interdisciplinary approach, development of AI/ML curricula requires input from many stakeholders if it is going to be successfully implemented. If we are going to provide educational programming that is realistic and meaningful, we must have the perspectives of all these stakeholders.

How does your research help people? Whom does it help?
We are trying to educate people that will be actively engaging with AI/ML-based technologies and considering the outputs of models when making decisions that impact patient lives. Therefore, this curriculum will help health care providers to provide data-driven, value-based, evidence-based care for the patients whom they are caring for.

What future direction will your research take?
Currently we are focused on increasing learners’ knowledge and understanding of concepts related to AI/ML in health care. Eventually, we would like to explore the impact that the curriculum is having on learner behaviors and, ultimately, patient outcomes. In addition, patients will need education about the role that AI/ML plays in their health care. This is another issue that I would like my team to address in the future.

What do you like to do when you aren’t doing research?
When I am not doing research, I am either teaching medical students and residents, or in the clinic seeing patients. I also enjoy traveling with my family, reading, and watching movies.