MR Image Reconstruction

MR image reconstruction aims to create high-quality images from incomplete or undersampled data, accelerating scan times and improving patient experience. Current research heavily utilizes deep learning, employing architectures like convolutional neural networks, transformers, and diffusion models, often incorporating k-space information and incorporating physics-based priors to improve accuracy and robustness. These advancements are crucial for improving the efficiency and diagnostic capabilities of MRI, particularly in applications like cardiac imaging and oncology where fast, high-resolution images are essential. A key focus is also on developing methods for reliable uncertainty quantification to ensure safe clinical translation.