MEDICAL IMAGE SYNTHESIS FOR RARE DISEASE USING AUGMENTATION VIA GANS

This research introduced GAN-based medical picture synthesis for uncommon disease data augmentation. High-quality synthetic medical images boost classification performance in the suggested  system.   Diffusion-based generative models, multimodal synthesis of imaging and clinical data, federated learning for crosshospital collaboration, and automated quality evaluation systems for synthetic data clinical reliability are future goals.

Y. Chen et al., “Conditional GANBased Medical Image Augmentation for Rare Disease Classification,” IEEE Transactions on Medical Imaging,  2023.   “MultiScale Adversarial Learning for Medical Image Synthesis,” IEEE Transactions on Neural Networks and Learning Systems, 2024.   Zhang et al., “Efficient GAN Architectures for HighResolution Medical Image Generation,” IEEE   Access,   2024.   R.     Patel     et     al.,     “AnatomyConstrained Generative Models for MRI Image Synthesis,” IEEE   Transactions   on      Biomedical Engineering,       2025.   H. Lee et al., “QualityAware GAN Training for Reliable Medical Image Augmentation,”IEEE   Transactions   on   AI,   2025.   A. Singh et al., “Lightweight Generative Models for Clinical Image Synthesis,” IEEE Internet   of      Things      Journal,       2026.   T. Nguyen et al., “Data Imbalance Handling in Medical   Image Classification Using Generative Models,” IEEE Access, 2023.   P.     Wang   et    al.,    “Adversarial     Training Strategies for Biomedical Image Generation,” IEEE   Transactions         on      AI, 2024.   K. Roy et al., “Perceptual Loss Optimisation for Medical Image Synthesis,” IEEE Access, 2025.   Brown et al., “Synthetic Data Generation for Healthcare AI Applications,” IEEE Access, 2026.   “Robust GAN Frameworks for Rare Disease Imaging,” IEEE Transactions on Neural Networks    and    Learning    Systems,     2024.   D. Sharma et al., “Evaluation Metrics for Medical Image Generation Models,” IEEE Access, 2025.