Abstract:Purpose: MRI represents an important diagnostic modality; however, its inherently slow acquisition process poses challenges in obtaining fully-sampled k-space data under motion. In the absence of fully-sampled acquisitions, serving as ground truths, training deep learning algorithms in a supervised manner to predict the underlying ground truth image becomes challenging. To address this limitation, self-supervised methods have emerged as a viable alternative, leveraging available subsampled k-space data to train deep neural networks for MRI reconstruction. Nevertheless, these approaches often fall short when compared to supervised methods. Methods: We propose Joint Supervised and Self-supervised Learning (JSSL), a novel training approach for deep learning-based MRI reconstruction algorithms aimed at enhancing reconstruction quality in cases where target datasets containing fully-sampled k-space measurements are unavailable. JSSL operates by simultaneously training a model in a self-supervised learning setting, using subsampled data from the target dataset(s), and in a supervised learning manner, utilizing datasets with fully-sampled k-space data, referred to as proxy datasets. We demonstrate JSSL's efficacy using subsampled prostate or cardiac MRI data as the target datasets, with fully-sampled brain and knee, or brain, knee and prostate k-space acquisitions, respectively, as proxy datasets. Results: Our results showcase substantial improvements over conventional self-supervised methods, validated using common image quality metrics. Furthermore, we provide theoretical motivations for JSSL and establish rule-of-thumb guidelines for training MRI reconstruction models. Conclusion: JSSL effectively enhances MRI reconstruction quality in scenarios where fully-sampled k-space data is not available, leveraging the strengths of supervised learning by incorporating proxy datasets.

SubZero: Subspace Zero-Shot MRI Reconstruction

Deep Learning for Under-Sampled Non-Cartesian ASL MRI Reconstruction

Zero-Shot Self-Supervised Learning for MRI Reconstruction

Zero-DeepSub: Zero-Shot Deep Subspace Reconstruction for Rapid Multiparametric Quantitative MRI Using 3D-QALAS

Zero‐DeepSub: Zero‐shot deep subspace reconstruction for rapid multiparametric quantitative MRI using 3D‐QALAS

NLCG-Net: A Model-Based Zero-Shot Learning Framework for Undersampled Quantitative MRI Reconstruction

Unsupervised MRI Reconstruction via Zero-Shot Learned Adversarial Transformers

Zero-Shot Image Denoising for High-Resolution Electron Microscopy

Improved Multi-Shot Diffusion-Weighted MRI with Zero-Shot Self-Supervised Learning Reconstruction

DFUSNN: zero-shot dual-domain fusion unsupervised neural network for parallel MRI reconstruction

Data-Consistent Non-Cartesian Deep Subspace Learning for Efficient Dynamic MR Image Reconstruction

Zero-shot Dynamic MRI Reconstruction with Global-to-local Diffusion Model

ZSDT: Zero-shot Domain Translation for Real-World Super-Resolution

Coordinate-Based Neural Representation Enabling Zero-Shot Learning for 3D Multiparametric Quantitative MRI

Self-supervised Learning for MRI Reconstruction with a Parallel Network Training Framework

CT Super Resolution via Zero Shot Learning

Towards Ultrafast MRI via Extreme k-Space Undersampling and Superresolution

JSSL: Joint Supervised and Self-supervised Learning for MRI Reconstruction

A Zero-Shot Method for 3D Medical Image Segmentation.

Accelerating Low-field MRI: Compressed Sensing and AI for fast noise-robust imaging