Search papers, labs, and topics across Lattice.
The paper introduces OccSAM-Bench, a new benchmark for evaluating the robustness of SAM-family segmentation models to surgical occlusions, using synthesized occlusions on polyp datasets. They propose a three-region evaluation protocol (full, visible-only, invisible) to decompose segmentation performance. Results reveal two distinct model archetypes: "Occluder-Aware" models that prioritize visible tissue and "Occluder-Agnostic" models that confidently predict into occluded regions, highlighting the need for task-specific model selection.
SAM models exhibit surprisingly divergent behaviors under occlusion, with some prioritizing visible tissue and others confidently hallucinating hidden anatomy.
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.
CMU ML