In an article published on 15 March in the European Journal of Physiology, Olivia Chevalier (Research Fellow) and Gérard Dubey (Professor of Sociology) at Institut Mines-Télécom Business School present a review of the literature on the contribution of AI in operating theatres.
The article analyses the impact of AI on the three phases of surgery: the preoperative, intraoperative and postoperative periods. The researchers examine the empirical evidence of AI’s clinical usefulness in operating theatres, as well as the ethical issues it raises.
Surgeons divided over AI in the operating theatre
The research team categorised the surgeons who took part in the survey as either enthusiasts, sceptics or cautious. The first group are those who are ready to fully integrate artificial intelligence into their medical practices. They feel that this tool far exceeds their expectations given the current implementations. The second group recognise the theoretical advances made by AI in operating theatres, but are concerned about the practicalities, particularly with regard to surgeons’ most complex tasks. The third group, the cautious ones, recognise the current limitations of AI. However, they remain optimistic and see these limitations as obstacles to be overcome by future innovations.
The article highlights this division within the medical profession by mentioning a few figures. These figures come from a previous study conducted by the German Surgical Association. According to the survey, 85% of the surgeons questioned believe that AI will have a positive impact on their field. However, only a third of them envisage AI playing a central role in the operating theatre. This demonstrates a clear interest in the technology. However, the majority (82%) are still reluctant to incorporate AI into hospital practices. This study highlights the need for a more in-depth analysis of practitioners’ perspectives and the real effects of incorporating AI into hospitals.
Systematic review of the literature using the PRISMA method.
The researchers used the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) method to produce this paper. This method ensures the transparency and thoroughness of the documents studied. The researchers focused on the PubMed, Scopus and IEEE Xplore databases between 2010 and 2023. The following keywords were used to refine the sample: ‘Artificial Intelligence’, ‘Machine Learning’, ‘Deep Learning’, ‘Surgical Procedures’, ‘Operative’ and ‘Clinical Decision-Making’. They also used the PICO method to define data inclusion and exclusion criteria. They therefore focused on:
- The population: patients undergoing surgical procedures in any field
- The intervention: AI technologies (including machine learning, deep learning, and robotic assistants) implemented in pre-, intra-, and postoperative care.
- The comparison: conventional surgical methods without AI assistance.
- Outcomes: changes in clinical outcomes, such as complication rates, efficiency indicators, surgeon performance, and patient safety.
AI results and limitations
This compilation of work makes it possible to consider the views of surgeons on the use of AI in their jobs. The variety of viewpoints presented provides future research and policy guidance to facilitate the integration of AI into operating theatres. For example, sceptics may require more empirical evidence from clinical trials to validate the benefits of AI in operating theatres. Conversely, enthusiasts may require specific training to enhance their understanding of AI. This will help to ensure that the tool is used correctly and not overestimated.
However, the tool is not without its faults. For example, artificial intelligence is not capable of handling anomalies that may occur during an operation. The human body is a complex and often unpredictable organism. It cannot be understood using deterministic scales. If we want AI to play a greater role in the operating theatre, its response time must be reduced as much as possible to avoid hindering operations that require speed. A legal and ethical framework also needs to be established to define the appropriate uses of AI. This will eventually enable AI to be effectively and safely integrated into operating theatres around the world.
