提出可控遮挡基准，评估医学分割模型在内窥镜遮挡下的表现。
Seeing Through the Tool: A Controlled Benchmark for Occlusion Robustness in Foundation Segmentation Models
- 构建合成遮挡数据集，模拟器械覆盖和切口两种遮挡类型。
- 三阶段评估揭示模型对可见与不可见区域的不同响应模式。
- 发现模型分两类：认遮挡者与无视遮挡者，选型需依临床目标定。
遮挡（目标结构被手术器械或重叠组织部分遮挡）是临床内窥镜中基础分割模型面临的关键挑战，但研究仍不充分。我们提出OccSAM-Bench，一个系统评估SAM系列模型在受控合成手术遮挡下的基准。框架在三个公开息肉数据集上，模拟两种遮挡类型（器械叠加、切口）及三种校准严重度。提出新颖的三区域评估协议，将分割性能分解为完整、仅可见、不可见目标三部分。该指标揭示了标准无模态评估所掩盖的行为，暴露两类模型范式：遮挡感知型（SAM、SAM 2、SAM 3、MedSAM3），优先处理可见组织并拒绝器械；遮挡无关型（MedSAM、MedSAM2），自信预测遮挡区域。SAM-Med2D既非其类，且在所有条件下表现均差。结果表明，遮挡鲁棒性并非架构通用，模型选择必须依据具体临床意图——是保守地分割可见组织，还是进行隐藏解剖结构的无模态推断。
原文摘要 · Abstract (English)
Occlusion, where target structures are partially hidden by surgical instruments or overlapping tissues, remains a critical yet underexplored challenge for foundation segmentation models in clinical endoscopy. We introduce OccSAM-Bench, a benchmark designed to systematically evaluate SAM-family models under controlled, synthesized surgical occlusion. Our framework simulates two occlusion types (i.e., surgical tool overlay and cutout) across three calibrated severity levels on three public polyp datasets. We propose a novel three-region evaluation protocol that decomposes segmentation performance into full, visible-only, and invisible targets. This metric exposes behaviors that standard amodal evaluation obscures, revealing two distinct model archetypes: Occluder-Aware models (SAM, SAM 2, SAM 3, MedSAM3), which prioritize visible tissue delineation and reject instruments, and Occluder-Agnostic models (MedSAM, MedSAM2), which confidently predict into occluded regions. SAM-Med2D aligns with neither and underperforms across all conditions. Ultimately, our results demonstrate that occlusion robustness is not uniform across architectures, and model selection must be driven by specific clinical intent-whether prioritizing conservative visible-tissue segmentation or the amodal inference of hidden anatomy.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。