EyePCR: A Comprehensive Benchmark for Fine-Grained Perception, Knowledge Comprehension and Clinical Reasoning in Ophthalmic Surgery
This work addresses the problem of evaluating and improving clinical reliability in surgical video understanding for researchers and practitioners in medical AI, though it is incremental as it adapts existing methods to a new domain-specific benchmark.
The authors tackled the lack of benchmarks for multimodal large language models in high-stakes surgical settings by developing EyePCR, a large-scale benchmark for ophthalmic surgery analysis, which includes over 210k VQAs and a medical knowledge graph, and their domain-adapted model achieved the highest accuracy in perception tasks and outperformed open-source models in comprehension and reasoning, rivaling commercial models like GPT-4.1.
MLLMs (Multimodal Large Language Models) have showcased remarkable capabilities, but their performance in high-stakes, domain-specific scenarios like surgical settings, remains largely under-explored. To address this gap, we develop \textbf{EyePCR}, a large-scale benchmark for ophthalmic surgery analysis, grounded in structured clinical knowledge to evaluate cognition across \textit{Perception}, \textit{Comprehension} and \textit{Reasoning}. EyePCR offers a richly annotated corpus with more than 210k VQAs, which cover 1048 fine-grained attributes for multi-view perception, medical knowledge graph of more than 25k triplets for comprehension, and four clinically grounded reasoning tasks. The rich annotations facilitate in-depth cognitive analysis, simulating how surgeons perceive visual cues and combine them with domain knowledge to make decisions, thus greatly improving models' cognitive ability. In particular, \textbf{EyePCR-MLLM}, a domain-adapted variant of Qwen2.5-VL-7B, achieves the highest accuracy on MCQs for \textit{Perception} among compared models and outperforms open-source models in \textit{Comprehension} and \textit{Reasoning}, rivalling commercial models like GPT-4.1. EyePCR reveals the limitations of existing MLLMs in surgical cognition and lays the foundation for benchmarking and enhancing clinical reliability of surgical video understanding models.