Faster than a speeding bullet: Like Superman, artificial intelligence is being taken up at a startling rate — more rapidly, even, than the personal computer and the internet were when those technologies emerged.
In medicine, physicians are adopting AI to help analyze data from diagnostic tests, formulate diagnoses and treatments, calculate disease risk, summarize patient histories, and communicate with patients and colleagues. AI is supporting health care administrators by making tasks and workflows more efficient. It’s assisting researchers in basic science discovery and in finding new drugs or new uses for existing drugs.
With this accelerating change, how can health care providers and scientists learn how to use AI tools, recognize their weaknesses and pitfalls, and use them to best benefit patients?
At Harvard Medical School, the Office for External Education (OEE) is developing a continuum of educational offerings — from learning the basics of AI to grappling with real challenges in medicine — so physicians, scientists, business leaders, and innovators can keep up.
AI offerings at HMS began with an executive education course in 2021. This spring, an online HMX short course on natural language processing introduced the fundamentals of this AI technology and how it can be used in health care. With interest in learning about the topic high, nearly 850 faculty, staff, and students at the School have accessed the materials.
OEE has woven aspects of AI through existing programs offered for medical educators through the Harvard Macy Institute and developed additional courses dedicated to AI, including a continuing education course for clinicians and another for corporate learners.
Applying AI to improve clinical medicine and health care
Most medical practitioners have not received formal training in artificial intelligence, said David Roberts, dean for external education. OEE’s new two-day live virtual course, AI in Clinical Medicine, explores AI models, strategies for use, and pitfalls so physicians will be equipped to use these tools to optimize their expertise — allowing them to make better informed diagnoses, formulate more effective treatment plans, and improve attentiveness to their patients. Learners discuss the ethical implications, challenges, and opportunities inherent in integrating AI into medical practice.
The course is directed by Maha Farhat, the Gilbert S. Omenn, MD ’65, PhD Associate Professor of Biomedical Informatics in the Blavatnik Institute at HMS; Samir Kendale, HMS assistant professor of anaesthesia at Beth Israel Deaconess Medical Center; and Isaac (Zak) Kohane, the Marion V. Nelson Professor of Biomedical Informatics at HMS.
A two-month online course called AI in Health Care: From Strategies to Implementation, meanwhile, prepares leaders in the medical and health sectors to address challenges in their organizations by studying the advantages and limitations of AI systems currently being used in the field. Learners take away a capstone project in which they identify an unmet need and pitch an AI solution to bring about transformative change in their organization, taking into consideration the tool’s benefits and its disadvantages, including its potential for bias and harm.
The teaching team includes Andrew Beam, assistant professor of epidemiology at the Harvard T.H. Chan School of Public Health, as well as instructors from four other leading companies and universities.
All courses keep patient safety, equity, and well-being at the forefront.
Ideally, by harnessing AI, patients can benefit from more personalized attention and physicians can enjoy greater well-being, Roberts said.
“We want to provide the resources through these training programs to use these tools and not lose the amazing humanity of medicine in the process,” he said.
The benefit of foresight
HMS was well-positioned to begin offering training in AI for health care ahead of the curve, with the founding of both OEE and the Department of Biomedical Informatics about 10 years ago.
This foresight allowed the School to have in place the expertise, faculty, and training programs needed to swiftly develop and launch continuing medical education programming in AI as more people realized that the technology will enable — or push — physicians, scientists, and educators to practice medicine, do research, and teach in totally different ways.
We want to provide the resources through these training programs to use these tools and not lose the amazing humanity of medicine in the process.
David Roberts
“We started thinking about AI programs several years ago, and we are fortunate because now when everyone wants to start learning about AI, we have the programs already built,” Roberts said.
Beyond the courses OEE is creating, Roberts said the synergy from the broad learning community is exciting as the courses draw “smart people from all over the world.”
“That’s where it gets really good,” he said.
Empowering doctors and patients
Roberts hopes that AI use enables doctors to do more of the things that they love: interacting one-on-one with and empowering their patients.
“Our goal is to reduce suffering that comes from disease, and AI can enable patients to find information, validate their concerns, and connect them with solutions,” he said.
This is just the beginning of how AI will help physicians and health care systems be more powerful than a locomotive in delivering high-quality health care to patients.
In the next 5 or 10 years the rate of change in AI will be exponential, making education even more critical, Roberts added. “The power of what is yet to come is going to be astonishing.”
