Abstract:Background: Dynamic magnetic resonance imaging (dMRI) plays an important role in cardiac perfusion and functional clinical exams. However, further applications are limited by the speed of data acquisition. Objective: A low-rank plus sparse decomposition approach is often introduced for reconstructing dynamic magnetic resonance imaging (dMRI) from highly under-sampling K-space data. In this paper, the reconstruction problem of DMR is transformed into a low-rank tensor plus sparse tensor recovery problem. Methods: A sequentially truncated higher-order singular value decomposition method is proposed to quickly approximate the low-rank tensor space structure and learn sparse components by adding a tensor kernel norm to the low-rank tensor and a l1 norm to the sparse tensor to constrain the two parts at the same time. The optimization problem is solved by using the iterative soft-thresholding algorithm; therefore, under the premise of ensuring the accuracy of the data, the amount of computation can be effectively reduced. Results: Compared with the state-of-the-art methods, the experimental results show that the proposed method can achieve better performance in terms of reconstruction speed and reconstruction quality on 3D and 4D dMRI datasets. Conclusion: The multidimensional MRI time series is represented by the tensor tool and decomposed into low rank tensor terms and sparse tensor terms. The low rank spatial structure is captured by the adaptive ST-HOSVD for fast approximation and the sparse component is constrained efficiently with a sparsity transform and l1 norm. The optimization problem is solved by an iterative soft-thresholding algorithm. Through extensive 3D and 4D dMRI experiments, it is demonstrated that our method can achieve superior reconstruction performance and efficiency compared with the other three state-of-theart methods reported in the literature.

Delta Scan: Reconstruction and Protocol Design for Ultra-fast Longitudinal MRI

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

Parallel Imaging and Convolutional Neural Network Combined Fast MR Image Reconstruction: Applications in Low-Latency Accelerated Real-Time Imaging.

Accelerating Longitudinal MRI using Prior Informed Latent Diffusion

PILOT: Physics-Informed Learned Optimized Trajectories for Accelerated MRI

DELTA-MRI: Direct deformation Estimation from LongiTudinally Acquired k-space data

Reconstruction Of Undersampled Big Dynamic Mri Data Using Non-Convex Low-Rank And Sparsity Constraints

High Speed Compressed Sensing Reconstruction in Dynamic Parallel MRI Using Augmented Lagrangian and Parallel Processing

Deep learning-based reconstruction of highly accelerated 3D MRI

End-to-End Sequential Sampling and Reconstruction for MRI

A Novel Reconstruction Approach Combining Global and Local Low-rank Constraints for Undersampled Dynamic MRI

Accelerating Dynamic MRI Reconstruction Using Adaptive Sequentially Truncated Higher-Order Singular Value Decomposition

Mri Reconstruction Algorithm Using Conjugate Gradient Method with Improved Line Search

Data- and Physics-driven Deep Learning Based Reconstruction for Fast MRI: Fundamentals and Methodologies

On Retrospective k-space Subsampling schemes For Deep MRI Reconstruction

Compressed sensing for longitudinal MRI: An adaptive-weighted approach

A Synchronized Compressed Sensing Scan-Reconstruction Scheme in Magnetic Resonance Imaging

Scan-Specific Unsupervised Highly Accelerated Non-Cartesian CEST Imaging Using Implicit Neural Representation and Explicit Sparse Prior

Fast Low Rank column-wise Compressive Sensing for Accelerated Dynamic MRI

High-Fidelity Accelerated MRI Reconstruction by Scan-Specific Fine-Tuning of Physics-Based Neural Networks