Artificial intelligence for improving intraoperative surgical care
Andrew P. Bain,Carla N. Holcomb,Herbert J. Zeh,Ganesh Sankaranarayanan,Bain, Andrew P.
DOI: https://doi.org/10.1007/s44186-024-00268-z
2024-06-20
Global Surgical Education - Journal of the Association for Surgical Education
Abstract:Ensuring patient safety and surgical quality in the operating room is critical in modern surgical care. Artificial intelligence (AI) technologies can contribute to these aims in three major domains: intraoperative clinical decision support, team process monitoring, evaluation, and feedback, and technical proficiency. Computer vision and deep learning techniques applied to intraoperative video can identify instruments and anatomy, segment and predict surgical steps, and evaluate technical performance. These developments present opportunities for intraoperative decision support that may obviate errors in perception or judgment, enhancing decision making. Similar techniques have been applied to intraoperative adverse events, identifying and stratifying bleeding events or inadvertent injuries. Furthermore, AI technologies can play a pivotal role in monitoring and evaluating teamwork and safety events. Novel comprehensive monitoring approaches such as the Surgical Control Tower and Operating Room Black Box integrate multiple data streams including audio-visual data of operating room events. Rapid analysis of team performance in safety protocols such as checklists and debriefs and quantitative data on communication, distraction, and error provide a fuller picture of the operating room environment. Successful implementation hinges on stakeholder engagement, education, and a focus on quality improvement. Lastly, the integration of AI into computer assisted surgical platforms, such as robotic surgery, will likely continue to improve technical proficiency. Limitations exist with regards to AI in the operating room. As computer vision techniques are refined, a thoughtful approach to intraoperative decision support must be taken to provide value and foster engagement. Standardization in data sharing and reporting methodologies is needed. An ethical framework for implementation of AI supported tools is critical. In conclusion, intraoperative AI can revolutionize decision support, safety monitoring, and overall quality. Responsible integration is paramount and collaborative decision making and continuous evaluation are crucial as surgical quality and safety is pursued.