Promptable Medical Image Segmentation Competition Dataset
收藏资源简介:
Promptable Medical Image Segmentation Competition Dataset是由全球24个机构合作创建的大规模医学图像分割数据集,涵盖了10种常见的成像模态,包括CT、MRI、PET等。数据集包含1,809,644个图像-掩码对,主要来源于公开的医学图像分割数据集,并经过预处理以确保标准化。数据集的创建旨在推动轻量级、高效的医学图像分割模型的开发,特别适用于临床实践中的资源受限环境。该数据集的应用领域广泛,旨在解决医学图像分割中的多样性和高效性问题,促进模型的临床应用。
Promptable Medical Image Segmentation Competition Dataset is a large-scale medical image segmentation dataset co-developed by 24 institutions worldwide. It encompasses 10 common imaging modalities including CT, MRI, PET and others. The dataset contains 1,809,644 image-mask pairs, which are primarily sourced from public medical image segmentation datasets and preprocessed to ensure standardization. This dataset is developed to promote the development of lightweight and efficient medical image segmentation models, and is particularly suited for resource-constrained clinical practice environments. With a wide range of application domains, it aims to address the challenges of diversity and efficiency in medical image segmentation and facilitate the clinical deployment of related models.

- 1Efficient MedSAMs: Segment Anything in Medical Images on Laptop研究机构 · 2024年



