AIMED25

Dr Anthony Chang

“If I have seen further, it is by standing on the shoulders of giants.”

                                                                                                                                       Isaac Newton

 

Beyond the aforementioned papers, books, and online videos, there are a myriad of other ways one can learn in the burgeoning domain of AI in medicine and healthcare leveraging the knowledge and wisdom of others.

Generative AI.

Use of large language models (LLMs) such as ChatGPT 4o and others like OpenEvidence can be good 24/7 virtual resources to learn not only AI and AI in healthcare but also clinical medicine and healthcare in general (see previous AIMed newsletter on the various LLMs in clinical medicine). AI pedagogy in medical education and clinical training entails learning about AI as well as learning using AI methodologies. In my experience, ChatGPT4o and now GPT-5 are very good at providing information about AI and its concepts and also for nuanced cases; these LLMs have been by far the most useful LLMs on a day-to-day basis for the myriad of tasks including differential diagnosis, justification letters, summarization of references, etc that I encounter in the clinical setting. In addition, medical students are using these LLMs to learn conventional subjects such as physiology and pharmacology (mainly text-based subjects but multimodal AI such as Google’s Gemini 1.5 or Anthropic’s Claude 3 will add to the learning capabilities for subjects like anatomy and histology). Current generative AI tools for visual content (text-to-picture) in medical education such as DALL-E 3 or Midjourney are not very mature, but my early experience with GPT-5 has been promising. All stakeholders interested in AI in healthcare can learn effectively and efficiently about AI from LLMs and also use these AI tools for their clinical practice and/or healthcare administration tasks. Unlike static resources, LLMs provide adaptive learning, adjusting explanations to a learner’s background and probing their understanding through question–answer exchanges. These generative AI tools can also summarize complex manuscripts, highlight key points from books, or translate highly technical material into accessible language for novices. Some of the LLMs, especially GPT4o or GPT-5 and Claude, can even code and explain each line for the beginner.

AI in Healthcare Virtual Learning Communities.

A more interactive approach to online learning has been the advent of virtual learning communities (VLCs) in which participants at all levels can post queries and make comments so everyone can learn from each other asynchronously. Unlike traditional classrooms or solitary study, a VLC leverages peer-to-peer exchange, mentorship, and real-time discussion across geographic and professional boundaries. The numerous other names for these virtual learning communities include: online learning community, digital learning network, virtual study group, learning circle, distance learning group, community of practice (or CoP), online knowledge exchange forum, or peer learning/knowledge sharing network (ACC- collaborative cognitive network?). An example of this virtual learning community is the AIMed25 WhatsApp community that started with a few ardent advocates of AI in health in the aftermath of AIMed25, and has now grown close to 650 participants from all over the world. Through these connections via WhatsApp or other platforms (such as Slack or WeChat) that can accommodate a myriad of documents and presentations, one can often find information and opinions that are considerably more valuable and insightful (and maybe even trusted) than what one can glean at random from the internet. Obviously, the value of communities like this is heavily dependent on the esprit de corps and expertise of its members that can ensure that knowledge sharing is not only vertical (from expert to novice) but also horizontal (across specialties and institutions) in this dynamic ecosystem of continuous learning.

AI in Healthcare Internship (Local Health Systems).

There are several opportunities for high school and college students to spend some time during summer months to focus on AI and related technologies in health. Health systems such as Stanford and Children’s Hospital of Orange County (now Rady Children's Health) offer such programs for rising high school and college students who have an interest in healthcare with an early focus on AI and other technologies. At Rady Children’s Health and Medical Innovation, Information, Investigation, and Intelligence (Mi4) summer internship program, the summer interns spend time with about a hundred mentors from healthcare and are encouraged to prepare and present abstracts for national and international meetings on AI in healthcare. When thoughtfully designed, such programs provide opportunities for students to shadow clinicians using AI-enabled tools, participate in data annotation projects, assist with basic literature reviews, or contribute to multidisciplinary team discussions that include computer scientists, ethicists, and healthcare professionals. These experiences not only demystify AI but also cultivate an appreciation for the complexities of medical data, regulatory considerations, and the collaborative nature of innovation in healthcare. To maximize impact, internships should incorporate structured mentorship, reflective journaling, and periodic debrief sessions where students critically connect their experiences to foundational AI concepts. By engaging in this way, young learners develop early AI literacy, an understanding of ethical and equity issues, and a sense of how technology can augment human care—skills and perspectives that will scaffold their future professional training.

AI in Healthcare Meetings (Local Schools or Health Systems).

Anyone can start a local AI in Healthcare meeting on a regular basis: a monthly convening of local interested parties is a good start. The monthly meeting can include an update of AI and AI in healthcare news as well as presentation of an AI in healthcare manuscript from journals; an occasional guest speaker can also provide an opportunity to introduce a topic. This is an effective way to gather the stakeholders from local colleges and universities as well as healthcare organizations to have an exchange of ideas and information on a regular basis. Even high school students have started these monthly meetings at their high schools. One can make the argument that such meetings should occasionally be in person to develop in person relationships amongst the stakeholders. An example of this monthly meeting is the virtual monthly Mi4 meetings that often have over 100 attendees. A monthly virtual meeting dedicated to artificial intelligence in healthcare can therefore serve as a cornerstone for sustained, community-wide learning across all levels of healthcare training and leadership. By providing a regular, structured forum, these sessions allow participants to stay current with rapidly evolving developments while fostering dialogue across disciplines and professional roles. Importantly, these meetings should be archived to create a digital knowledge repository accessible to those unable to attend in real time. By committing to a predictable monthly rhythm, participants develop a habit of continuous learning, and over time, the meeting becomes a community of practice—a place where premedical students, clinicians, educators, and healthcare executives alike can exchange insights, challenge assumptions, and collectively shape the future of AI integration in healthcare.

AI in Healthcare Courses (Local University or Health Professional School).

For the students in colleges and universities, enrollment in artificial intelligence in healthcare courses can be a good introduction. Very few health professional schools have such a course, but more students are requesting to have an AI in healthcare course than ever before. A few schools, however, do have an elective of AI in healthcare. The deans of these schools will need to be cognizant of both the technological trends as well as the growing sentiment of the students for formal education in this domain. Indoctrinating artificial intelligence into a healthcare professional curriculum requires moving beyond isolated lectures toward a longitudinal and embedded approach that threads AI literacy across the stages of training. Rather than treating AI as an optional elective or a technical curiosity, it should be presented as a core competency— much like anatomy, ethics, or pharmacology— that all healthcare professionals must understand to practice effectively in the 21st century. The integration process is most effective when scaffolded, starting with basic awareness, progressing to critical appraisal, and culminating in strategic and ethical leadership. By weaving AI into multiple touchpoints—lectures, simulations, clinical rotations, capstone projects, and continuing medical education—curricula can ensure that healthcare professionals not only gain technical literacy but also develop the critical judgment, ethical reasoning, and collaborative mindset necessary to harness AI responsibly in clinical practice and healthcare systems.

The education of AI in healthcare for everyone will be a dedicated track at AIMed25 this year.

Artificial Intelligence in Medicine (AIMed) is the longest running meeting (inaugurated in 2013) focused on artificial intelligence in medicine and healthcare. The meeting is usually attended by well over 1,000 clinicians (from over 50 subspecialties) as well as healthcare leaders, data scientists, students and trainees, and entrepreneurs and investors from all over the world. The 3-day meeting covers a broad range of topics directly or indirectly related to artificial intelligence in medicine and healthcare such as generative AI, agentic AI, large language models, cybersecurity, and intelligent extended reality with three special tracks this year: AI in pediatrics and neonatology, AI in health professional education, and AI and mental health of clinicians and patients. Several special features of this meeting of AI in health include: breakfast workshops focused on hot topics (as determined by our attendees as attendees vote on topics they like to hear), afternoon subspecialty breakout sessions (for over 20 different subspecialties or domain areas), abstract competition with scholarships for accepted scholars, and a special one-day American Board of AI in Medicine (ABAIM) course that has become very popular with our attendees. This year, there will be a special Chief AI Officer agenda during AIMed25.

See you there!