Scholarlink

<p>High-quality segmentation datasets are essential for advancing AI applications in medical imaging. However, it is challenging to generate such datasets for highly variable and complex organs like the colon. We introduce a dataset of 435 human colons, segmented from Computed Tomography Colonography (CTC) obtained from the publicly available The Cancer Imaging Archive (TCIA). Each scan includes a mask of the whole colon, including collapsed segments and the fluid, and a mask of only the gas-filled parts of the colon. The colon segmentation accuracy has been clinically validated by an expert abdominal radiologist. This is the first open-access dataset of segmented colons derived from CTC. This resource enables population-scale radiologic studies, supports the development of AI-based image analysis tools, and facilitates the creation of anatomically accurate digital models and simulators, both virtual and physical.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Clinically validated dataset of 435 human colons segmented from CT colonography

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年1月14日

查看全文
  • Martina Finocchiaro,
  • Ronja Stern,
  • Rikke Vilhelmsborg,
  • Abraham George Smith,
  • Jens Petersen,
  • Kristoffer Cold,
  • Lars Konge,
  • Kenny Erleben,
  • Melanie Ganz

摘要

High-quality segmentation datasets are essential for advancing AI applications in medical imaging. However, it is challenging to generate such datasets for highly variable and complex organs like the colon. We introduce a dataset of 435 human colons, segmented from Computed Tomography Colonography (CTC) obtained from the publicly available The Cancer Imaging Archive (TCIA). Each scan includes a mask of the whole colon, including collapsed segments and the fluid, and a mask of only the gas-filled parts of the colon. The colon segmentation accuracy has been clinically validated by an expert abdominal radiologist. This is the first open-access dataset of segmented colons derived from CTC. This resource enables population-scale radiologic studies, supports the development of AI-based image analysis tools, and facilitates the creation of anatomically accurate digital models and simulators, both virtual and physical.