Dear Editor,
In the dynamic landscape of medical education, the fusion of artificial intelligence (AI) and simulation technologies has ushered in a new era of immersive learning experiences.[1] Several innovations use AI technology available in the public domain [Table 1]. Among these innovations, the Conversational Generative Pre-trained Transformer (ChatGPT), an AI-powered conversational agent developed by OpenAI, is leading the charge in revolutionising medical training. ChatGPT is enhancing prebriefing, scenario creation, and debriefing and redefining the educational paradigm for aspiring healthcare professionals through its seamless integration into simulation zones.[2]
Table 1.
Artificial intelligence chatbots for medical education and training simulation
| Chatbot | Description | Potential in medical education/training | Strengths | Weaknesses | Pricing | |||||
|---|---|---|---|---|---|---|---|---|---|---|
| Bard (Google) | Large language model trained on a massive dataset of text and code | Can simulate patient interviews, answer medical questions and provide feedback on clinical decisions | Strong factual knowledge, ability to reason and generate different responses, customisable | Not specifically designed for medical education, may not understand complex medical concepts | Free for non-commercial use | |||||
| ChatGPT (OpenAI) | Large language model is known for its creative text generation capabilities | Can roleplay as patients presenting various symptoms, provide emotional responses and engage in open-ended conversations | Engaging and creative responses, ability to adapt to different conversation styles | Not as factually accurate as other models, it may struggle with complex medical terminology | Free for personal use, paid plans for developers | |||||
| LaMDA (Google AI) | Conversational AI focuses on realistic and engaging dialogue | Can simulate patient interactions with diverse backgrounds and personalities, practise communication skills and address ethical dilemmas | Human-like conversation style, ability to learn and adapt to user preferences | Limited medical knowledge, may not be able to handle complex medical scenarios | Closed access, not available to the public | |||||
| Mitsuku | Award winning chatbot known for its wit and humour | Can simulate patients with mental health conditions, practise empathy and active listening skills, and explore the impact of emotions on medical interactions | Witty and engaging personality, ability to improvise and respond creatively | Not specifically designed for medical education, limited knowledge of medical terminology | Free for personal use, paid plans for developers | |||||
| Florence Nightingale (Simulations in Nursing Education) | AI-powered virtual patient simulator for nursing students | Can simulate a wide range of medical conditions, practise clinical assessments and interventions, and receive feedback on performance | Specific to nursing education, realistic patient simulations with customisable scenarios | Limited interaction beyond clinical tasks, not designed for open-ended conversations | Paid access to educational institutions |
Bard Google is now ‘Gemini’ by Google. Mitsuku is now Kuki AI developed by Steve Worswick. AI=artificial intelligence, ChatGPT=Conversational Generative Pre-trained Transformer developed by OpenAI, LaMDA=Language Model for Dialogue Applications by Google AI
Prebriefing sessions, essential for setting learning objectives and establishing psychological safety, are enriched by ChatGPT’s ability to engage students in interactive discussions about upcoming scenarios. ChatGPT provides guidance, clarifies expectations and fosters a sense of readiness as students prepare to enter simulated clinical environments, ensuring optimal immersion and learning outcomes.[3]
ChatGPT is a versatile tool for educators to design diverse and realistic patient encounters in scenario creation. Educators can collaborate with ChatGPT through natural language interaction to script-nuanced scenarios that challenge students’ clinical reasoning and communication skills. Whether simulating a routine patient consultation or a high-stakes medical emergency, ChatGPT adapts its responses based on student actions, creating dynamic and authentic learning experiences.
During simulations, ChatGPT assumes the role of virtual patients, seamlessly integrating into the scenario to provide lifelike interactions. Students engage in clinical assessments, formulate diagnoses and execute treatment plans while conversing with ChatGPT, which responds with realistic feedback and cues. The adaptive nature of ChatGPT allows for tailored responses that mirror real-world patient interactions, enhancing the authenticity and efficacy of the simulation experience.[4]
ChatGPT facilitates debriefing sessions by providing structured feedback, prompting reflective questions and guiding students through self-assessment exercises. By leveraging AI-driven analytics, ChatGPT identifies learning gaps and areas for improvement, empowering students to address weaknesses and refine their clinical skills iteratively.
Moreover, ChatGPT’s role extends beyond individual simulation sessions to encompass longitudinal learning trajectories. ChatGPT offers insights into student progress and competence development over time by storing and analysing data from multiple interactions. Educators can track performance trends, identify patterns of growth and tailor learning interventions accordingly, fostering continuous improvement and mastery of clinical competencies. Kundra and Senthilnathan,[5] in a recent editorial, highlighted how machine learning–powered simulations in anaesthesia help doctors make better decisions for patients with complex medical conditions.
In conclusion, ChatGPT’s integration into simulation zones represents a paradigm shift in medical education, where AI-driven technologies enhance every facet of the learning journey. From prebriefing to scenario creation, from immersive simulations to debriefing, ChatGPT enriches the educational experience, empowering students to become skilled, empathetic and resilient healthcare practitioners. As the frontier of medical training continues to evolve, ChatGPT stands as a beacon of innovation, shaping the future of healthcare education one conversation at a time.
Financial support and sponsorship
Nil.
Conflicts of interest
There are no conflicts of interest.
REFERENCES
- 1.Garg R. Simulation, artificial intelligence and technology. Indian J Anaesth. 2024;68:5–7. doi: 10.4103/ija.ija_1268_23. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Eysenbach G. The role of ChatGPT, generative language models, and artificial intelligence in medical education: A conversation with ChatGPT and a call for papers. JMIR Med Educ. 2023;9:e46885. doi: 10.2196/46885. doi: 10.2196/46885. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Johnson-Bey S, Mateas M, Wardrip-Fruin N. Toward using ChatGPT to generate theme-relevant simulated storyworlds. AIIDE Workshop on Experimental Artificial Intelligence in Games. 2023;3626 Available from: https://ceur-ws.org/Vol-3626/short5.pdf. [Last accessed on 2024 April 13] [Google Scholar]
- 4.Schmidt A, Elagroudy P, Draxler F, Kreuter F, Welsch R. Simulating the human in HCD with ChatGPT: Redesigning interaction design with AI. Interactions. 2024;31:24–31. [Google Scholar]
- 5.Kundra P, Senthilnathan M. Amalgamation of artificial intelligence and simulation in anaesthesia training: Much-needed future endeavour. Indian J Anaesth. 2024;68:8–10. doi: 10.4103/ija.ija_1264_23. [DOI] [PMC free article] [PubMed] [Google Scholar]
