A systematic scoping review protocol to summarise and appraise the use of artificial intelligence in the analysis of digital videos of invasive general surgical procedures

• Published in: International journal of surgery protocols Vol. 27; no. 3; pp. 118 - 121
• Main Authors: King, Anni, Fowler, George, Macefield, Rhiannon C., Quek, Fang-Fang, Walker, Hamish, Thomas, Charlie, Markar, Sheraz, Blazeby, Jane M., Blencowe, Natalie S.
• Format: Journal Article
• Language: English
• Published: Hagerstown, MD Lippincott Williams & Wilkins 01.12.2023
• Subjects: Protocol
• ISSN: 2468-3574; 2468-3574
• DOI: 10.1097/SP9.0000000000000012

Background: Intraoperative video recordings are a valuable addition to operative written documentation. However, the review of these videos often requires surgical expertise and takes considerable time. While a large amount of work has been undertaken to understand the role of artificial intelligence (AI) in healthcare more generally, the application of these techniques to automate the analysis of surgical videos is currently unclear. In this systematic scoping review, we sought to give a contemporary overview of the use of AI research in the analysis of digital videos of invasive general surgical procedures. We will describe and summarise the study characteristics, purpose of the applications and stage of development, to ascertain how these techniques might be applied in future research and to identify gaps in current knowledge (e.g. uncertainties about the study methods). Methods: Systematic searches will be conducted in OVID Medline and Embase, using terms related to ‘artificial intelligence’, ‘surgery’ and ‘video’ to identify all potentially relevant studies published since 1st January 2012. All primary studies where AI has been applied to the analysis of videos (recorded by conventional digital cameras or laparoscopic or robotic-assisted technology) of general surgical procedures will be included. Data extraction will include study characteristics, governance, details of video datasets and AI models, measures of accuracy, validation and any reported limitations. Ethics and dissemination: No ethical approval is required as primary data will not be collected. The results will be disseminated at relevant conferences, on social media and published in a peer-reviewed journal.