LUMID: Large-scale Unlabled Medical Imaging Dataset for Unsupervised and Self-supervised Learning

LUMID is a large-scale, unlabeled collection of over 2 million medical images spanning multiple imaging modalities, including CT scans, X-rays, MRIs, and more. This dataset has been meticulously curated from publicly available medical imaging repositories, addressing the critical challenge of limited scale in existing public datasets and the inaccessibility of high-quality private datasets. The primary motivation behind creating this dataset is to empower the medical imaging community with a resource suited for developing and training advanced deep learning models. By enabling the use of unsupervised and self-supervised learning approaches, this dataset facilitates the learning of rich, transferable representations that can significantly enhance performance across various medical imaging tasks, including classification, segmentation, and anomaly detection. Key Features: 1) Diversity: Comprising images from multiple modalities and a wide range of medical imaging scenarios. 2) Scalability: A dataset of unprecedented size, providing a robust foundation for training deep neural networks. 3) Versatility: Specifically designed for unsupervised and self-supervised learning methods, fostering innovation in representation learning for medical imaging. 4) Open Access: Built entirely from public datasets, ensuring transparency and reproducibility. This dataset is intended to serve as a cornerstone for advancing research in medical AI, fostering the development of models capable of generalizing across diverse imaging types and clinical conditions.