Medical Segmentation Decathlon (MSD) datasets
收藏资源简介:
医学分割十项全能(MSD)数据集是由纪念斯隆-凯特琳癌症中心等多家机构合作创建的大型注释医学图像数据集。该数据集包含2633个三维图像,涵盖多种感兴趣的解剖结构、多种模态和来源,用于支持语义分割算法的开发和评估。数据集通过开放源许可证提供,旨在通过全面的基准测试客观评估通用分割方法,并为研究领域提供开放和免费的医学图像数据。数据集的应用领域广泛,包括肿瘤、心脏、肝脏等器官的分割,旨在解决临床治疗规划和肿瘤体积测量等问题。
The Medical Segmentation Decathlon (MSD) dataset is a large-scale annotated medical imaging dataset collaboratively developed by multiple institutions including Memorial Sloan Kettering Cancer Center. This dataset comprises 2633 3D images, covering a diverse range of anatomical structures of interest, imaging modalities and data sources, and is designed to support the development and evaluation of semantic segmentation algorithms. Distributed under an open-source license, the dataset aims to objectively evaluate general-purpose segmentation methods through comprehensive benchmarking, while providing open and free medical imaging data for the global research community. It has a wide range of application scenarios, including segmentation of targets such as tumors, heart, liver and other anatomical structures, with the goal of addressing practical clinical issues including treatment planning and tumor volume quantification.




