Files
Abstract
The 2024 ImageCLEFmedical GANs task Controlling the Quality of Synthetic Medical Images created via GANsis in its second edition. It comprises two sub-tasks which address the security and privacy concerns related to personal medical image data in the context of generating and using synthetic images in different real-life scenarios. The first sub-task is an extension of the task presented in the previous edition, focusing on examining the hypothesis that generative models (e.g., GANs, Diffusion Models) generate medical images containing certain “fingerprints” of the original images used for network training. The second sub-task, new this year, explores the hypothesis that generative models imprint unique fingerprints on generated images. The focus is on understanding whether different generative models or architectures leave discernible signatures within the synthetic images they produce. Ground truth data was made available to the participants. This paper presents the overview of systems and runs submitted by describing the datasets, the evaluation metrics, and discussing the methods proposed by the participating teams and their results.