Search papers, labs, and topics across Lattice.
The paper introduces Cholec80-port, a new high-fidelity dataset for trocar port segmentation in laparoscopic videos, addressing the lack of geometrically consistent port annotations in existing surgical datasets. They define a rigorous standard operating procedure (SOP) for port segmentation that excludes the central lumen, ensuring geometric consistency. Experiments demonstrate that training with Cholec80-port and SOP-consistent annotations significantly improves the robustness of segmentation models across different datasets compared to using larger, geometrically inconsistent datasets.
Erroneous surgical port annotations are holding back geometry-based surgical AI, but a new geometrically consistent dataset fixes this.
Trocar ports are camera-fixed, pseudo-static structures that can persistently occlude laparoscopic views and attract disproportionate feature points due to specular, textured surfaces. This makes ports particularly detrimental to geometry-based downstream pipelines such as image stitching, 3D reconstruction, and visual SLAM, where dynamic or non-anatomical outliers degrade alignment and tracking stability. Despite this practical importance, explicit port labels are rare in public surgical datasets, and existing annotations often violate geometric consistency by masking the central lumen (opening), even when anatomical regions are visible through it. We present Cholec80-port, a high-fidelity trocar port segmentation dataset derived from Cholec80, together with a rigorous standard operating procedure (SOP) that defines a port-sleeve mask excluding the central opening. We additionally cleanse and unify existing public datasets under the same SOP. Experiments demonstrate that geometrically consistent annotations substantially improve cross-dataset robustness beyond what dataset size alone provides.
