EUVIMEDEUVIMED entdecken
Crossref

609 The Use of Artificial Intelligence to Correctly Identify Key Anatomical Structures During Laparoscopic Appendectomies

S Meher, G Janakan

British Journal of Surgery · 10.07.2026 · 113 (Supplement_6)

DOI: 10.1093/bjs/znag063.611 ↗

Kurz gefasst

Fachartikel in der Zeitschrift British Journal of Surgery, online erschienen am 10.07.2026. Verfasst von S Meher und 1 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Aim

This study evaluated the diagnostic accuracy of multimodal artificial intelligence (AI) models in identifying key anatomical structures during laparoscopic appendectomy and explored their potential usefulness in supporting trainee surgeons and improving intra-operative decision-making.

Method

A cross-sectional image-recognition analysis was conducted using clinical images obtained from open-access sources and peer-reviewed surgical publications, including those behind cookie walls. Laparoscopic images represented a full range of appendiceal and bowel pathology: normal appendices, inflamed and necrotic appendices, gangrenous tissue, mucocele, and Meckel’s diverticulum or small bowel. Two clinicians independently verified anatomical identification to establish a reference standard. Three multimodal AI platforms—ChatGPT (OpenAI), Gemini (Google), and Copilot (Microsoft)—were assessed using a standard diagnostic question evaluating their ability to correctly identify the appendix and surrounding structures. Sensitivity and specificity were calculated for each model.

Results

ChatGPT demonstrated the highest diagnostic accuracy, achieving a sensitivity of 83% and specificity of 80% followed by Gemini with a sensitivity of 55% and specificity of 60%. Copilot demonstrated <10% sensitivity and was unable to complete specificity testing due to image-blocking of sensitive surgical content. Platforms performed better when anatomy was clearly visualised and accuracy declined with complex pathology such as severe inflammation or necrosis.

Conclusions

AI-assisted anatomical recognition in laparoscopic surgery appears increasingly feasible. Future implementation will require safeguards for patient privacy and image security, particularly for cloud-based systems. At the current pace of development, integration of real-time AI image recognition into the operating room is likely within the next five years, with potential to support surgeons and trainees.

Zitieren

APA

Meher, S., & Janakan, G. (2026). 609 The Use of Artificial Intelligence to Correctly Identify Key Anatomical Structures During Laparoscopic Appendectomies. British Journal of Surgery, 113(Supplement_6). https://doi.org/10.1093/bjs/znag063.611

Vancouver

Meher S, Janakan G. 609 The Use of Artificial Intelligence to Correctly Identify Key Anatomical Structures During Laparoscopic Appendectomies. British Journal of Surgery. 2026;113(Supplement_6). doi:10.1093/bjs/znag063.611

BibTeX
@article{Meher2026,
  author = {Meher, S and Janakan, G},
  title = {609 The Use of Artificial Intelligence to Correctly Identify Key Anatomical Structures During Laparoscopic Appendectomies},
  journal = {British Journal of Surgery},
  year = {2026},
  volume = {113},
  number = {Supplement_6},
  doi = {10.1093/bjs/znag063.611}
}
Kostenlos starten →