Web-Based Medical Image Repositories: Characterization and Searching

Medical images are essential for diagnosing and treating diseases, as well as for developing Artificial Intelligence tools for these purposes. However, obtaining high-quality datasets is expensive and their availability may be limited. Fortunately, the growth of online, open-access repositories provides researchers with a valuable resource for medical image data. The challenge lies in the lack of a unified system to classify these repositories, making it difficult to find the right dataset for a specific research project. This work addresses the aforementioned issue by proposing a systematic analysis of the existing repositories and defines key characteristics for comparison. Researchers can then leverage this work to identify datasets that precisely match their needs, considering factors like disease focus, imaging modality, and access restrictions.