Abstract:

Introduction:
In 1950, Alan Turing first mentioned the use of computers in simulating intelligent behaviour for critical thinking. Artificial intelligence (AI) has documented surgical applications in forecasting operation durations, outcomes and postoperative complications. Surgical education and training (SET) processes combine didactic learning, acquiring supervised clinical skills and hands-on operative training. The COVID pandemic has rapidly changed the way education is delivered worldwide with AI gaining popularity in surgery. Chat Generative Pretrained Transformer (ChatGPT) is a publicly available novel natural language processing AI chatbot. As the surgical arena necessitates more practical training compared to others, the aim is to review the role of ChatGPT in SET as a future training tool.

Methodology:
Literature review done from 2020–2024, using scientific electronic databases and relevant keywords to gather the pertinent data.

Results:
The role of ChatGPT at different levels of SET:
Undergraduate level medical students can benefit from ChatGPT for their surgical exam preparations in creating topic summaries and multiple-choice questions (MCQ). ChatGPT-4 scored 76% in surgical board exam with the ability to understand complex surgical information.
Resident level-theory of surgical principles and surgical techniques learning can be augmented by ChatGPT, saving time as it generates procedure summaries. Surgical residents learn from experienced surgeons regarding clinical decision-making on the need for surgery, operative procedure, preoperative and postoperative optimisation. Natural language processing deals with pattern recognition of data has the potential to aid in pre-operative surgical decision making and post-operative care. Virtual reality integrated ChatGPT could potentially enable an engaging interactive simulation training experience. Capabilities in generating intraoperative surgical steps through accepting images, analysing important anatomical landmarks and providing personalised feedback could optimise residents’ learning in surgical skills. This could be valuable for intraoperative navigation, currently under exploration.
Postgraduate level chatbots automate repetitive administrative tasks like discharge summaries, operative notes and patient letters while saving clinicians time. It cannot analyse video messages in real time surgery yet. Hence, it has not been tried in operative rooms. It provides strategies for handling of difficult patients using active listening, empathy, open honest communication and setting boundaries. However, human emotions should not be underestimated in patient-surgeon relationships. Chatbots can produce evidence-based scientific information updates for continuous professional development.

Limitations include errors due to artificial hallucinations, information governance and medicolegal concerns. The Bletchley declaration for artificial intelligence safety was signed in November 2023 to regulate AI risks. Comprehensive AI chatbot validity evaluation tools should be developed.

Conclusion:
ChatGPT should be considered as an additional hand, not a replacement, in surgery to enhance human capabilities with the implementation of adequate safety measures for SET at different levels. Intraoperative SET needs experienced surgeons as human teachers. Artificial intelligence has not overruled human intelligence in surgery as of yet.

Keywords:

Surgical Education, Artificial Intelligence, ChatGPT, Medical Students.