[2603.04340] Balancing Fidelity, Utility, and Privacy in Synthetic Cardiac MRI Generation: A Comparative Study

Computer Science > Computer Vision and Pattern Recognition arXiv:2603.04340 (cs) [Submitted on 4 Mar 2026] Title:Balancing Fidelity, Utility, and Privacy in Synthetic Cardiac MRI Generation: A Comparative Study Authors:Madhura Edirisooriya, Dasuni Kawya, Ishan Kumarasinghe, Isuri Devindi, Mary M. Maleckar, Roshan Ragel, Isuru Nawinne, Vajira Thambawita View a PDF of the paper titled Balancing Fidelity, Utility, and Privacy in Synthetic Cardiac MRI Generation: A Comparative Study, by Madhura Edirisooriya and 7 other authors View PDF Abstract:Deep learning in cardiac MRI (CMR) is fundamentally constrained by both data scarcity and privacy regulations. This study systematically benchmarks three generative architectures: Denoising Diffusion Probabilistic Models (DDPM), Latent Diffusion Models (LDM), and Flow Matching (FM) for synthetic CMR generation. Utilizing a two-stage pipeline where anatomical masks condition image synthesis, we evaluate generated data across three critical axes: fidelity, utility, and privacy. Our results show that diffusion-based models, particularly DDPM, provide the most effective balance between downstream segmentation utility, image fidelity, and privacy preservation under limited-data conditions, while FM demonstrates promising privacy characteristics with slightly lower task-level performance. These findings quantify the trade-offs between cross-domain generalization and patient confidentiality, establishing a framework for safe and effective synt...