SUREON: A Benchmark and Vision-Language-Model for Surgical Reasoning

TL;DR

SUREON introduces a large-scale surgical video QA dataset and models that enable AI to better understand surgical scenes, reasoning about intent, safety, and future steps, surpassing general models in accuracy.

Contribution

The paper presents SUREON, a novel dataset and models for surgical reasoning, leveraging expert narrations to train AI for complex surgical question answering.

Findings

SUREON dataset contains 206.8k QA pairs from 134.7K clips across 170 procedures.

Models outperform larger general-domain models, achieving over 84% accuracy on the benchmark.

Qualitative analysis shows models can infer operative intent and reasoning from visual data.

Abstract

Surgeons don't just see -- they interpret. When an expert observes a surgical scene, they understand not only what instrument is being used, but why it was chosen, what risk it poses, and what comes next. Current surgical AI cannot answer such questions, largely because training data that explicitly encodes surgical reasoning is immensely difficult to annotate at scale. Yet surgical video lectures already contain exactly this -- explanations of intent, rationale, and anticipation, narrated by experts for the purpose of teaching. Though inherently noisy and unstructured, these narrations encode the reasoning that surgical AI currently lacks. We introduce SUREON, a large-scale video QA dataset that systematically harvests this training signal from surgical academic videos. SUREON defines 12 question categories covering safety assessment, decision rationale, and forecasting, and uses a multi-agent pipeline to extract and structure supervision at scale. Across 134.7K clips and 170 procedure types, SUREON yields 206.8k QA pairs and an expert-validated benchmark of 354 examples. To evaluate the extent to which this supervision translates to surgical reasoning ability, we introduce two models: SureonVLM, a vision-language model adapted through supervised fine-tuning, and SureonVLM-R1, a reasoning model trained with Group Relative Policy Optimization. Both models can answer complex questions about surgery and substantially outperform larger general-domain models, exceeding 84% accuracy on the SUREON benchmark while outperforming general-domain models on standard surgical perception tasks. Qualitative analysis of SureonVLM-R1 reveals explicit reasoning behavior, such as inferring operative intent from visual context.