Deep Learning for Medical Image Enhancement: A Systematic Review of Architectures, Imaging Modalities, and Open Challenges

Abstract

Medical image enhancement plays a pivotal role in improving diagnostic accuracy across various healthcare applications. In recent years, deep learning (DL) techniques have emerged as powerful tools for enhancing medical images, outperforming traditional image processing algorithms. This systematic review provides a comprehensive overview of recent advancements in deep learning architectures applied to medical image enhancement, including Convolutional Neural Networks (CNNs), Generative Adversarial Networks (GANs), U-Net variants, and Vision Transformers (ViTs). We analyze their application across major imaging modalities such as Magnetic Resonance Imaging (MRI), Computed Tomography (CT), Ultrasound, and X-ray. Furthermore, this review highlights key performance metrics, publicly available datasets, and critical open challenges including data scarcity, model interpretability, and clinical deployment. Finally, potential future directions are outlined to bridge the gap between algorithmic developments and practical clinical workflow.