Swarm Learning performs comparably to centralized learning for laparoscopic appendicitis grading.
Privacy-Preserving Surgical Video Analysis with Swarm Learning - Results from a Multinational Appendectomy Cohort.
O L Saldanha … F R Kolbinger
NEJM AI · 2026
Background
Progress in artificial intelligence (artificial intelligence)-based analysis of surgical videos has been constrained by reliance on manual frame-level annotations rather than patient-level outcomes.
Methods
We evaluated our pipeline using a dataset of 397 laparoscopic appendectomy recordings from six international centers for two patient-level staging tasks: (1) laparoscopic grading of appendicitis and appendiceal perforation detection; and (2) histopathologic inflammation grading.
n = 397 laparoscopic appendectomy
AUROC: 0.795±0.092
Results
AUROC: 0.795±0.092
Swarm Learning achieved an AUROC of 0.795 for accurately grading laparoscopic appendicitis.
n = 397 laparoscopic appendectomy
“
Conclusion · 1 of 3
Weakly supervised deep learning enables the prediction of patient-level labels directly from surgical video data.
Conclusion · 2 of 3
Swarm Learning facilitates privacy-preserving multicenter collaboration and achieves performance on par with centralized learning, highlighting its potential for advancing clinically relevant, collaborative artificial intelligence development in surgical video analysis.