Nicolas Padoy

A single AI model can outperform specialized counterparts in recognizing surgical phases across multiple centers, challenging the need for tailored approaches.

SPIRIT achieves superior surgical action triplet recognition by leveraging relational modeling, outperforming traditional methods across diverse surgical environments.

Performance leakage in federated surgical AI can exceed 80%, but GEN-Guard effectively corrects these failures, enhancing model robustness across institutions.

Gaze-following in the operating room isn't just about watching eyes; it's a surprisingly effective lens for understanding clinical roles, surgical phases, and team communication, boosting performance by over 30% in some cases.

Surgical VQA gets a major upgrade: SurgTEMP's hierarchical visual memory and competency-based training leapfrog existing models in understanding complex, time-sensitive surgical procedures.