Educators from University of Michigan’s Medical School Push for AI Training in the Medical School Curriculum

A group of educators and researchers from University of Michigan’s Medical School called for integrating AI and machine learning into the medical school curriculum in a recent Cell Reports Medicine article. They claimed that AI and its usage is relevant to nearly all areas of clinical practice and that current medical school graduates are left “under prepared” to interface effectively with these technologies. 

Two of the article’s authors, Erkin Ötleş, a machine learning researcher and current medical and Ph.D. Student at the University of Michigan, and Jim Woolliscroft, a former Michigan Medical School dean, followed up on these ideas in an interview in Stat Magazine.

Ötleş expressed concern about medical providers who use AI systems without the understanding needed to ask questions and validate the output, as this can lead to unchecked system errors or biases that cause harm. “We’re going to be at a point where we’re not going to be able to catch up and be able to call out the technology defects or flaws,” Ötleş said. “Without being armed with that set of foundational knowledge into how these things work, we’re going to be at a disadvantage.”

Ötleş and Woolliscroft also described the current training in AI and machine learning in medical school as predominantly student-driven. Interested students are mostly left to independently seek out additional degrees, courses, or electives as there is currently no systemic approach to teach medical students about AI in an integrated and meaningful way. “Medical students don’t know about this stuff, and they need to see it as basic as pharmacology and physiology. Already, machine learning algorithms, and more generally AI, are essentially ubiquitous,” Woolliscroft said.

In Stat Magazine, the authors proposed a “spiral curriculum” for AI, which would introduce and reintroduce AI topics to medical students routinely and within different contexts. Students would start with the basics, then circle back later to learn about AI alongside other specialized knowledge. While neither Ötleş nor Woolliscroft suggested that medical students need to become programmers, they did say that integrating AI into the curriculum will empower medical students to ask validating questions about how the AI works and about the data underlying it. 

“So, when they’re on radiology, they can ask: So this mammogram interpretation, what was it based on? Did it include women from, say, Egypt that have a lot more inflammatory breast cancer? It didn’t. Oh, OK. Well, here in Michigan, we have a lot of people from the Middle East. So is this going to be applicable to this population or not? As they get into all of these different things, they’ll have a foundation that they can plug in these specific examples to fill out the flesh of those bones that have been laid.” Woolliscroft said.

Survey Finds Law Students Hesitant About Generative AI Technology

The use of generative AI technology is growing within the legal industry. In a recent LexisNexis survey, over half of lawyers polled reported using the technology for research (59 percent) and improving efficiency (54 percent). Significant proportions also used it for drafting documents (45 percent) and writing emails (34 percent). 

Surprisingly, however, the survey found that law students were the least likely group to report using the technology. Only nine percent of the responding law students said that they use AI currently in law school and a quarter said that they plan to use it in their future legal work. 

So why aren’t law students jumping on the AI bandwagon? Serena Wellen of LexisNexis took a deep dive into the data to find out. We’ve summarized her findings below: 

• Accuracy. Some of the law students noted that generative AI is not reliable. Research findings may be inaccurate, false, or based on unreliable data that it presents as fact, even including false citations. 

• Academic Integrity. Students fear that the use of generative AI could encourage academic dishonesty as the proper use of the tools has not been well defined at many law schools.

• Innovative Thinking. Students note that learning and practicing the law requires critical and innovative thinking and they believe that the use of generative AI would discourage their refinement of this skill-set.  

• Fear of Replacement. Some students fear that generative AI will overtake entry-level legal positions that offer essential learning opportunities. 

Ultimately, while generative AI will likely play an increasingly important role in legal work, students are correct in expressing their hesitation. Law schools will need to develop guidelines for proper use of the technology.

Doctors Seek Additional Training as Technology and Big Data Converge with Patient Care

Classes in data analytics, artificial intelligence, and technology may sound like the course load of an MBA student, but a recently released report shows the immense value of this subject matter for medical students as well. The Stanford Medicine Trends in Health 2020 report entitled “The Rise of the Data Driven Physician” describes a future for physicians where data and technology are increasingly intermingled with effective patient care.

The report, which includes a survey of physicians (n=523), residents (n=133), and current medical students (n=77), analyzes how trends in medicine impact those on the front lines of patient care.  The health care sector, the report claims, is undergoing “seismic shifts, fueled by a maturing digital health market, new health laws that accelerate data sharing, and regulatory traction for artificial intelligence in medicine.”  And the report’s findings show that while there is acceptance and even enthusiasm around the benefits of data and technology among providers, there is also a real gap that exists between the training that physicians, residents, and students receive, and the demands of the evolving profession. Lloyd Minor, MD and Dean of the Stanford University School of Medicine says, “We’ve found that current and future physicians are not only open to new technologies, but are actively seeking training in subjects such as data science to enhance care for their patients. We are encouraged by these findings and the opportunity they present to improve patient outcomes. At the same time, we must be clear-eyed about the challenges that may stymie progress.”

The respondents appear keenly aware of the changes occurring within medicine and both physicians and residents say that they expect about a quarter of their current duties to be automated using technology over the next 20 years. Students predict that, on average, 30 percent of their duties will be automated. Further, just under half of the physicians surveyed (47 percent) and about three-quarters of the students surveyed (73 percent) say that they are currently seeking out training to better prepare themselves for innovations in health care. Among physicians who said they are currently attending training, the most popular subjects are genetic counseling (38 percent), artificial intelligence (34 percent), and population health management (31 percent). Students seeking additional training courses are taking advanced statistics and data science (44 percent), population health management (36 percent), and genetic counseling (30 percent) at the highest rates.

When asked about which innovations have the most potential to transform health care in the next five years, physicians as well as students and residents (grouped) were both most likely to respond with personalized medicine, followed by telemedicine. Both groups also see potential in artificial intelligence, wearable health monitoring devices, and genetic screening.

When the groups were asked about how helpful their education has been in preparing them for new technologies in healthcare, just under 20 percent of both students and residents (18 percent) and physicians (19 percent) responded “very helpful.” Current students and residents were more positive overall, with 58 percent responding that their education was “somewhat helpful.” Only 23 percent responded negatively. Among physicians, 36 percent found their education somewhat helpful, and the remaining 44 percent replied negatively.

The most notable gap, perhaps, is the one highlighted below, which showcases the difference between the perceived benefits of medical innovation for patients and how prepared the provider is in implementing that innovation. The two charts, students and residents (grouped) and physicians, show the group that agreed an innovation will be “very beneficial to future patients” compared to the group that said they feel “very prepared to use the innovation in practice.” Some of the gaps are marked. For example, over half of students and residents (55 percent) and physicians (51 percent) believe that personalized medicine will be very beneficial to future patients, yet only five percent of students and residents and 11 percent of physicians feel prepared to deliver it. Personalized medicine showed the largest gap between perceived benefit and preparedness for both groups, while virtual reality showed the smallest. Both groups feel very prepared to work with electronic health records, but there were lower levels of perceived benefit to the patient.

Gap Between Perceived Benefits to Patients and Provider Preparedness

This report demonstrates the changing nature of the medical profession, its current and continuing intersection with technology and big data, as well as the need to provide opportunities and training to medical providers so that they can use these innovations to improve patient care. Prospective and current medical students will want to carefully consider how they are using their time prior to and during medical school. It may be beneficial for them to spend time speaking with physicians and residents to gauge what technologies and research are driving change. Taking classes prior to medical school, in statistics and data modeling, as well as technology and AI may have a positive impact on their ability to bridge the gap between present demands and the traditional medical school curriculum.