Table 2: Summary of AI and digital health technologies, applications, and evidence in medical education.
|
Technology Domain |
Key Applications in Medical Education |
Evidence Summary |
Reported Limitations |
Key References |
|
Large language models (ChatGPT, GPT-4, Gemini, Claude) |
Examination-item generation; virtual tutoring; clinical-case simulation; literature synthesis; personalized study plans |
GPT-4 accuracy 81.1% on USMLE; item quality comparable to human-generated; potential bias in clinical vignettes |
Hallucination of content; algorithmic bias; over-reliance; data-privacy concerns |
[12,19,20,22] |
|
Virtual reality / augmented reality |
Anatomy education; surgical simulation; clinical-scenario training; procedural skills; communication training |
Improved knowledge retention, procedural skills, and engagement versus traditional methods; effective for anatomy knowledge gains |
High cost; infrastructure needs; limited long-term outcome data; motion sickness in some learners |
[14,15,23,24] |
|
Simulation-based training |
High-fidelity manikins; VR simulation; standardized patients; gamified training; telemedicine simulation |
Enhanced skill acquisition and error reduction; effective repeated practice; improved confidence and clinical judgment |
Resource-intensive; requires trained facilitators; fidelity-effectiveness relationship unclear |
[2,5] |
|
Telemedicine education |
Telehealth-competency training; remote encounters; virtual consultations; digital-workflow education |
83% of students desire telemedicine training; feasible simulation-based assessment; growing competency frameworks |
Limited standardized curricula; assessment tools still developing; digital-infrastructure gaps |
[16,55,56] |
|
Adaptive and e-learning platforms |
Personalized pathways; spaced repetition; performance-based content adjustment; learning analytics |
Spaced digital education improved retention; adaptive platforms enhanced mastery rates by up to 30% |
Algorithm-transparency issues; one-size-fits-all design risks; variable engagement |
[18,21] |
|
Mobile health and wearables |
Skill-practice apps; physiological monitoring during learning; point-of-care references; collaborative learning |
Increased accessibility and flexibility; real-time feedback; potential for continuous assessment |
Screen fatigue; distraction; data-privacy concerns; limited educational evidence |
[6,17] |