CHAOS数据集

CHAOS（Combined Healthy Abdominal Organ Segmentation）数据集是一个公开的多模态医学图像分割数据集，专注于健康腹部器官（肝脏、左肾、右肾和脾脏）的分割任务。它包含 CT 和 MR 图像，旨在促进跨模态学习和医学图像分析算法的开发。
CHAOS 数据集的主要目标是填补多模态医学图像分割的空白。它提供了配对的 CT 和 MR 图像，支持研究人员探索如何融合两种模态的信息以提高分割精度。数据集特别关注健康个体的器官分割，挑战算法在不同模态下的鲁棒性。
数据集包含约 40 名健康受试者的扫描数据，每名受试者至少包含一个 CT 扫描和多个 MR 扫描序列（如 T1 加权和 T2 加权）。所有图像均经过医学专家的精细标注，标注质量经过多次审阅和验证。

CHAOS (Combined Healthy Abdominal Organ Segmentation) dataset is a publicly available multimodal medical image segmentation dataset focused on segmenting healthy abdominal organs, namely the liver, left kidney, right kidney, and spleen. It comprises CT and MR images, with the goal of advancing cross-modal learning and the development of medical image analysis algorithms.

The primary goal of the CHAOS dataset is to address the gap in multimodal medical image segmentation research. It offers paired CT and MR images, allowing researchers to investigate methods for fusing information from both modalities to enhance segmentation accuracy. The dataset specifically targets organ segmentation in healthy individuals, posing challenges for algorithms to achieve robustness across different imaging modalities.

The dataset contains scanning data from roughly 40 healthy subjects. Each subject has at least one CT scan and multiple MR scan sequences, such as T1-weighted and T2-weighted sequences. All images have been meticulously annotated by medical professionals, and the annotation quality has undergone multiple rounds of review and validation.

搜集汇总

背景与挑战

背景概述

CHAOS数据集是一个公开的多模态医学图像分割数据集，专注于健康腹部器官（肝脏、左肾、右肾和脾脏）的分割任务，包含CT和MR图像以支持跨模态学习。它基于约40名健康受试者的扫描数据，所有图像均经过医学专家的精细标注。