Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

SegCol Challenge: Semantic Segmentation for Tools and Fold Edges in Colonoscopy data

Xinwei Ju, Rema Daher, Razvan Caramalau, Baoru Huang, Danail Stoyanov, Francisco Vasconcelos

TL;DR

SegCol Challenge tackles colonoscopy navigation by enabling semantic segmentation of fold edges and endoscopic tools to support depth perception and localization. It leverages an EndoMapper-based dataset of 96 videos with pixel-level labels for fold edges and three instrument classes, organized into train/validation/unlabeled/test splits and 40-frame batches. The framework defines two tasks—model design for segmentation and active learning to select informative frames—evaluated with a multi-metric suite (Dice, AP, OIS, ODS, CLDice) and careful zero-handling to accommodate thin contours and dense masks. This benchmark aims to drive depth-aware localization and potential 3D reconstruction in colonoscopy, with practical implications for improving CRC screening navigation systems.

Abstract

Colorectal cancer (CRC) remains a leading cause of cancer-related deaths worldwide, with polyp removal being an effective early screening method. However, navigating the colon for thorough polyp detection poses significant challenges. To advance camera navigation in colonoscopy, we propose the Semantic Segmentation for Tools and Fold Edges in Colonoscopy (SegCol) Challenge. This challenge introduces a dataset from the EndoMapper repository, featuring manually annotated, pixel-level semantic labels for colon folds and endoscopic tools across selected frames from 96 colonoscopy videos. By providing fold edges as anatomical landmarks and depth discontinuity information from both fold and tool labels, the dataset is aimed to improve depth perception and localization methods. Hosted as part of the Endovis Challenge at MICCAI 2024, SegCol aims to drive innovation in colonoscopy navigation systems. Details are available at https://www.synapse.org/Synapse:syn54124209/wiki/626563, and code resources at https://github.com/surgical-vision/segcol_challenge .

SegCol Challenge: Semantic Segmentation for Tools and Fold Edges in Colonoscopy data

TL;DR

SegCol Challenge tackles colonoscopy navigation by enabling semantic segmentation of fold edges and endoscopic tools to support depth perception and localization. It leverages an EndoMapper-based dataset of 96 videos with pixel-level labels for fold edges and three instrument classes, organized into train/validation/unlabeled/test splits and 40-frame batches. The framework defines two tasks—model design for segmentation and active learning to select informative frames—evaluated with a multi-metric suite (Dice, AP, OIS, ODS, CLDice) and careful zero-handling to accommodate thin contours and dense masks. This benchmark aims to drive depth-aware localization and potential 3D reconstruction in colonoscopy, with practical implications for improving CRC screening navigation systems.

Abstract

Colorectal cancer (CRC) remains a leading cause of cancer-related deaths worldwide, with polyp removal being an effective early screening method. However, navigating the colon for thorough polyp detection poses significant challenges. To advance camera navigation in colonoscopy, we propose the Semantic Segmentation for Tools and Fold Edges in Colonoscopy (SegCol) Challenge. This challenge introduces a dataset from the EndoMapper repository, featuring manually annotated, pixel-level semantic labels for colon folds and endoscopic tools across selected frames from 96 colonoscopy videos. By providing fold edges as anatomical landmarks and depth discontinuity information from both fold and tool labels, the dataset is aimed to improve depth perception and localization methods. Hosted as part of the Endovis Challenge at MICCAI 2024, SegCol aims to drive innovation in colonoscopy navigation systems. Details are available at https://www.synapse.org/Synapse:syn54124209/wiki/626563, and code resources at https://github.com/surgical-vision/segcol_challenge .

Paper Structure

This paper contains 8 sections, 1 figure, 1 table.

Table of Contents

  1. Introduction
  2. Dataset details
  3. Dataset Sampling
  4. Dataset annotation
  5. SegCol Challenge
  6. Baseline model and Evaluation
  7. Metrics:
  8. Implementation:

Figures (1)

  • Figure 1: SegCol Evaluation Workflow. The flowchart places data preparation on the left block, with Task 1 and Task 2 represented by the upper and lower blocks on the right.