Video based assessment with AI in open surgery
British Journal of Surgery · 10.08.2026 · 113 (Supplement_9)
DOI: 10.1093/bjs/znag093.101 ↗Kurz gefasst
Fachartikel in der Zeitschrift British Journal of Surgery, online erschienen am 10.08.2026. Verfasst von Annie Brange und 5 weiteren Autor:innen.
Abstract (Originalsprache)
Introduction
The technical ability and precision of surgeons impact surgical complications in patients. Surgical training is competence based, and residents often perceive insufficient feedback from their supervisors. In laparoscopic and robotic assisted surgical procedures, video analysis with AI has shown promising results for feedback of surgical performance. Many procedures are still conducted with open technique. The aim of this study was to investigate the feasibility of multicamera recording and video-based artificial intelligence analysis of instrument and hand movements of open surgical procedures.
Methods
Patients under the age of 15 undergoing orchidopexy were included. A camera rig with two cameras oriented cranially and caudally were mounted on the operating table approximately 1,1m over the patient. Video was captured of the operating field with a fixed focus set on the skin and subcutaneous level. Ethical consent was received. A Segment Anything Model 2 (SAM 2, Meta, Menlo Park, CA) for propagation-based tracking and YOLOv26 for detection-based monitoring (YOLOv26, Ultralytics, Frederick, MD, USA) was trained to detect and track surgical instruments from the recorded videos.
Results
Five procedures produced over 15 000 frames feasible for further analysis. The AI software was able to identify, detect and follow the surgical instruments in the video.
Discussion
AI video analysis of open surgical procedures is feasible and a promising method for objective assessment of surgery. Its potential for assessment and use as an objective feedback instrument of surgical technique is still unclear. Present challenges include developing a user-friendly feedback interface.
Zitieren
APA
Brange, A., Lindberg, F., Lundström, S., Stenmark, M., Tran, P., & Hagelsteen, K. (2026). Video based assessment with AI in open surgery. British Journal of Surgery, 113(Supplement_9). https://doi.org/10.1093/bjs/znag093.101
Vancouver
Brange A, Lindberg F, Lundström S, Stenmark M, Tran P, Hagelsteen K. Video based assessment with AI in open surgery. British Journal of Surgery. 2026;113(Supplement_9). doi:10.1093/bjs/znag093.101
BibTeX
@article{Brange2026,
author = {Brange, Annie and Lindberg, Felix and Lundström, Svante and Stenmark, Maj and Tran, Phan-Kiet and Hagelsteen, Kristine},
title = {Video based assessment with AI in open surgery},
journal = {British Journal of Surgery},
year = {2026},
volume = {113},
number = {Supplement_9},
doi = {10.1093/bjs/znag093.101}
}
EUVIMED