Compressive sensing MR imaging based on adaptive tight frame and reference image

In this paper, we present a fast iterative magnetic resonance imaging (MRI) reconstruction algorithm taking advantage of the prevailing GPGPU programming paradigm. In clinical environment, MRI reconstruction is usually performed via fast Fourier transform (FFT). However, imaging artifacts (i.e. signal loss) resulting from susceptibility-induced magnetic field inhomogeneities degrade the quality of reconstructed images. These artifacts must be addressed using accurate modeling of the physics of the system coupled with iterative reconstruction. We have developed a reconstruction algorithm with improved image quality at the expense of computation time and hence an implementation on GPUs achieving significant speedup. In this work, we extend our previous work on GPU implementation by adding several new features. First, we enable Sensitivity Encoding for Fast MRI (SENSE) reconstruction (from data acquired using a multi-receiver coil array) which can reduce the acquisition time. Besides, we have implemented a GPU-based total variation regularization in our SENSE reconstruction framework. In this paper, we describe the different optimizations employed from levels of algorithm, program code structures, and specific architecture performance tuning, featuring both our MRI reconstruction algorithm and GPU hardware specifics. Results show that the current GPU implementation produces accurate image estimates while significantly accelerating the reconstruction.

This paper introduces an efficient algorithm for magnetic resonance (MR) image reconstruction. The proposed method minimizes a linear combination of nonlocal total variation and least-square data-fitting term to reconstruct the MR images from undersampled k-space data. The nonlocal total variation is taken as the L 1-regularization functional and solved using Split Bregman iteration. The proposed algorithm is compared with previous methods in terms of the reconstruction accuracy and computational complexity. The comparison results demonstrate the superiority of the proposed algorithm for compressed MR image reconstruction.

Deep learning-based magnetic resonance imaging (MRI) reconstruction obtains extremely high reconstruction performance by training with large amounts of data. However, the acquisition of fully sampled MRI data is expensive and time-consuming, making it difficult to obtain large amounts of fully sampled data. This not only poses a challenge for models that require more fully sampled data for training but also limits the generalization ability of network models. To address the above problems, we propose a semi-supervised MRI reconstruction algorithm based on migration learning by analyzing the characteristics of under-sampled MRI reconstruction. To address the lack of data scarcity and generalization in the reconstruction process. The model fuses shallow features extracted from the convolutional layer and deep features obtained from the Swin Transformer Block (STB) to greatly improve the reconstruction capability of the network. However, the process of distillation cannot predict the goodness of the knowledge transferred, which hinders the performance of the semi-supervised algorithm. We further propose the use of privilege loss as a way to improve the distillation of useful knowledge. Experimental results show that student networks trained by this algorithm are able to rival supervised teacher networks and outperform unsupervised network models in brain data reconstruction.KeywordsMagnetic resonance imagingSwin transformerKnowledge distillationPrivileged information

An approach to Magnetic Resonance (MR) image reconstruction from undersampled data is proposed. Undersampling artifacts are removed using an iterative thresholding algorithm applied to nonlinearly transformed image block arrays. Each block array is transformed using kernel principal component analysis where the contribution of each image block to the transform depends in a nonlinear fashion on the distance to other image blocks. Elimination of undersampling artifacts is achieved by conventional principal component analysis in the nonlinear transform domain, projection onto the main components and back-mapping into the image domain. Iterative image reconstruction is performed by interleaving the proposed undersampling artifact removal step and gradient updates enforcing consistency with acquired k-space data. The algorithm is evaluated using retrospectively undersampled MR cardiac cine data and compared to k-t SPARSE-SENSE, block matching with spatial Fourier filtering and k-t ℓ1-SPIRiT reconstruction. Evaluation of image quality and root-mean-squared-error (RMSE) reveal improved image reconstruction for up to 8-fold undersampled data with the proposed approach relative to k-t SPARSE-SENSE, block matching with spatial Fourier filtering and k-t ℓ1-SPIRiT. In conclusion, block matching and kernel methods can be used for effective removal of undersampling artifacts in MR image reconstruction and outperform methods using standard compressed sensing and ℓ1-regularized parallel imaging methods.

In order to accelerate the imaging process of MRI (Magnetic Resonance Imaging), CS-MRI (Compressed Sensing Magnetic Resonance Imaging) utilizes the prior information of MR images and reconstructs under-sampled MR images by solving an optimization problem through an iterative algorithm, and it is time consuming. With the emergence of deep learning technology, more and more optimization problems can be solved through well-designed networks and large amounts of training data. Therefore, DL-Based (Deep Learning-Based) methods have attracted much attention. However, most of the current work simply treats MRI reconstruction as an image-to-image task, so there is still a lot of room for improvement in how to better integrate MRI information. This paper proposes a new DL-based method, which reconstructs the MR images in both k-space and image domain using several sub-networks and fuses information among different sub-networks. We also used down-sampling and up-sampling pairs to reduce the computational complexity. Our proposed architecture obtained promising experimental results on Calgary-Campinas Brain MRI datasets.

The deep learning method is an efficient solution for improving the quality of undersampled magnetic resonance (MR) image reconstruction while reducing lengthy data acquisition. Most deep learning methods neglect the mutual constraints between the real and imaginary components of complex-valued k-space data. In this paper, a new complex-valued convolutional neural network, namely, Dense-U-Dense Net (DUD-Net), is proposed to interpolate the undersampled k-space data and reconstruct MR images. The proposed network comprises dense layers, U-Net, and other dense layers in sequence. The dense layers are used to simulate the mutual constraints between real and imaginary components, and U-Net performs feature sparsity and interpolation estimation for k-space data. Two MRI datasets were used to evaluate the proposed method: brain magnitude-only MR images and knee complex-valued k-space data. Several operations were conducted for data preprocessing. First, the complex-valued MR images were synthesized by phase modulation on magnitude-only images. Second, a radial trajectory based on the golden angle was used for k-space undersampling, whereby a reversible normalization method was proposed to balance the distribution of positive and negative values in k-space data. The optimal performance of DUD-Net was demonstrated based on a quantitative evaluation of inter-method and intra-method comparisons. When compared with other methods, significant improvements were achieved, PSNRs were increased by 10.78 and 5.74dB, whereas RMSEs were decreased by 71.53% and 30.31% for magnitude and phase image, respectively. It is concluded that DUD-Net significantly improves the performance of MR image reconstruction.

Regularized iterative reconstruction algorithms for imaging inverse problems require selection of appropriate regularization parameter values. We focus on the challenging problem of tuning regularization parameters for nonlinear algorithms for the case of additive (possibly complex) Gaussian noise. Generalized cross-validation (GCV) and (weighted) mean-squared error (MSE) approaches (based on Steinfs Unbiased Risk Estimate. SURE) need the Jacobian matrix of the nonlinear reconstruction operator (representative of the iterative algorithm) with respect to the data. We derive the desired Jacobian matrix for two types of nonlinear iterative algorithms: a fast variant of the standard iterative reweighted least-squares method and the contemporary split-Bregman algorithm, both of which can accommodate a wide variety of analysis- and synthesis-type regularizers. The proposed approach iteratively computes two weighted SURE-type measures: Predicted-SURE and Projected-SURE (that require knowledge of noise variance Ð2), and GCV (that does not need Ð2) for these algorithms. We apply the methods to image restoration and to magnetic resonance image (MRI) reconstruction using total variation (TV) and an analysis-type .1-regularization. We demonstrate through simulations and experiments with real data that minimizing Predicted-SURE and Projected-SURE consistently lead to near-MSE-optimal reconstructions. We also observed that minimizing GCV yields reconstruction results that are near-MSE-optimal for image restoration and slightly suboptimal for MRI. Theoretical derivations in this work related to Jacobian matrix evaluations can be extended, in principle, to other types of regularizers and reconstruction algorithms.

Super-resolution reconstruction of single anisotropic 3D MR images using residual convolutional neural network

Exploiting rank deficiency and transform domain sparsity for MR image reconstruction

Radial encoding is often performed for fast data acquisition in magnetic resonance imaging (MRI). Therefore, the gridding method, which is a kind of algorithm for Fast Fourier Transform, is widely used in fast MRI reconstruction. In this paper, an improved gridding method has been utilized to 2D MRI reconstruction, which precomputes an 1D finely sampled table of a high-accuracy interpolator and then uses separable bilinear interpolation by look-up table. Experimental comparisons are made by assessing the performance of the direct reconstruction, the conventional KB gridding and the proposed method. The experimental results show that the proposed method provides high accuracy, fast computation, and dramatic memory saving.

Magnetic Resonance Imaging (MRI) reconstruction is computationally heavy from under sampled data, and centralized data sharing within deep learning models was met with privacy concerns. We therefore propose FL-Swarm MRCM, a novel federated learning framework that integrates FedDyn dynamic regularization, a swarm-optimized generative adversarial network (SwarmGAN), and a structure-aware cross-entropy loss to enhance cross-site MRI reconstruction without sharing raw data. The framework avoids client drift, locally adapts hyper-parameters using Particle Swarm Optimization, and preserves anatomic fidelity. Evaluations on fastMRI, BraTS-2020, and OASIS datasets under non-IID partitions show that FL-Swarm MRCM improves reconstruction quality, achieving PSNR = 29.78 dB and SSIM = 0.984, outscoring FL-MR and FL-MRCM baselines. The federated framework for adversarial training proposed here enables reproducible, privacy-preserving, and strongly multi-institutional MRI reconstruction with better cross-site generalization for clinical use.

Objectives:The objective of our study is to determine whether the addition of3-dimensional (3D) MRI to standard MRI sequences is comparable to 3D CT scanevaluation of glenoid and humeral bone loss in glenohumeral instability.There is increasing understanding of the importance of the glenoid, as wellas humeral bone loss in outcomes and surgical decision making in patientswith shoulder instability. Although magnetic resonance imaging (MRI) isperformed as standard of care for soft tissue evaluation of shoulderinstability, CT scan remains the gold standard in bony measurements.Standard MRI evaluation has been shown to inaccurately measure bone loss.However, additional CT scans result in increased financial, time, andionizing radiation cost to the patient.Methods:Eighteen patients who presented with glenohumeral instability wereprospectively enrolled and received both MRI and CT within 1 week of eachother. The MRI included an additional sequence (VIBE) which underwentpost-processing for reformations (Figure 1a). The addition of a VIBEprotocol, on average is an addition 4 to 4.5minutes in the scanner. CT dataalso underwent 3D post-processing (Figure 1b) and therefore each patient hadfour imaging modalities (2D CT, 2D MRI, 3D CT reformats, and 3D MRIreformats). Each sequence underwent the following measurements from twoseparate reviewers: glenoid defect, glenoid defect percent, humeral defect,humeral defect percentage (Figure 2a and 2b), and evaluation of glenoidtrack and version. Paired t-tests were used to assess differences betweenimaging modalities and chi squared for glenoid track. Intra-observer andinter-observer reliability were evaluated. Bland-Altman tests were alsoperformed to assess the agreement between CT and MRI. In addition, wedetermined cost of each imaging modality at our single institution.Results:3D MRI measurements for glenoid and humeral bone loss measurements werecomparable to 3D CT (Table 1). There were no significant differences forglenoid defect size and percentage, or humeral defect size and percentage(P>0.05) (Table 2). Bland-Altman analysis demonstrated strong agreementwith small measurement errors for 3D CT and 3D MRI percentage glenoid boneloss. There was also no difference in evaluation for determining on vs offtrack between any of the imaging modalities. Inter- and intra-raterreliability was good to excellent for all CT and MRI measurements (r >=0.7).Conclusions:3D MRI measurements for bone loss in glenohumeral instability throughutilization of VIBE sequence were equivalent to 3D CT. At our institution,the costs of MRI with 3D reconstruction was 1.67x cheaper than MRI and CTwith 3D reconstructions. 3D MRI may be a useful adjuvant to standard MRIsequences to allow concurrent soft tissue and accurate assessment of glenoidand humeral bone loss in glenohumeral instability.Figure 1a.(3D MRI glenoid reformation) and lb (3D CT glenoid reformation):Glenoid measurement for 3D reconstruction of MRI and CT usingsagittal en face of the glenoid. A circle was placed over theinferior portion of the glenoid using margins as a guide for thecircle. The glenoid height line (blue) runs from the supraglenoidtubercle to the medial aspect of the inferior base. Glenoid boneloss size (orange) and circle diameter (green) was quantified withglenoid height line marking the center of the circle. Percentglenoid bone loss was determined as [(glenoid bone loss size/circlediameter)* 100]. Glenoid track was determined through equation[(0.83 * circle diameter) - glenoid bone loss size]Figure 2a.(2D CT) and 2b (3D CT): 2a: Measurement of Hill Sachs maximal defecton axial 2D CT recorded in mm. 2b: Measurement of humeral defectpercentage from the 3D reformats using sagittal projection with bestfit circle. A line (blue) was placed to determine maximal humeralhead height (A) then perpendicular line (green) to measure residualhumeral head (B) and therefore bone defect (A-B)(orange). Percentbone loss was determined through [(A-B)/A)*100]Table 1:Mean 2D and 3D CT and MRI measurementsTable 2:Paired t-test Glenoid and Humeral Bone Loss Measurements.

DiffGAN: An adversarial diffusion model with local transformer for MRI reconstruction

In magnetic resonance imaging (MRI), low-resolution (LR) images often hamper clinical diagnosis and research due to constraints in imaging conditions and technology limitations. Recent studies in super-resolution (SR) reconstruction of multi-contrast MRI have shown promise by leveraging the complementary information from different MRI contrasts. However, existing multi-contrast MRI SR techniques face several challenges: 1) a lack of pre-alignment precision can result in distorted reconstructions; 2) prevailing transformer network structures, with their smaller windows (e.g., 8 × 8), struggle to effectively capture long-range dependencies and lack the ability to interact between different windows; and 3) current methods are limited to fixed integer scaling (e.g., 2 ×, 3 ×, 4 ×), which limits flexibility and increases complexity in training and storage. To address these challenges, we propose a novel arbitrary-scale SR network for multi-contrast MRI. Specifically, our approach compensates for spatial misalignment between modalities through deformable registration module and employs permuted cross-attention transformer in MR images. In addition, we introduce a ref-scale ensemble implicit attention module that better integrates high-frequency information from reference images and enables arbitrary-scale upsampling. Extensive experiments on two publicly available MRI datasets validate the superiority of our method in multi-contrast MRI SR, demonstrating its significant potential in clinical applications.

To determine the association between osseous injuries and short-term clinical outcome in patients with anterior cruciate ligament (ACL) tear. The retrospective study was performed with institutional review board approval, and the requirement to obtain informed consent was waived. The study group consisted of 114 patients (57 male and 57 female patients with a mean age of 26.1 and 25.1 years, respectively) with ACL tear who underwent magnetic resonance (MR) imaging and ACL reconstruction surgery and who filled out International Knee Documentation Committee (IKDC) knee evaluation questionnaires before and 1 year after surgery. All MR images were independently reviewed by two radiologists to determine the presence of a cortical depression fracture on each surface of the knee joint. Bone marrow edema volume was quantified by using segmentation software. Correlation coefficients were used to determine the association between bone marrow edema volume and IKDC score. A multivariate analysis model was used to compare IKDC scores in patients without fracture, patients with a single fracture, and patients with multiple fractures. There was no significant association between total bone marrow edema volume and preoperative or postoperative IKDC score (P = .32 and P = .91, respectively). The mean preoperative and postoperative IKDC scores were 53.4 ± 18.6 and 93.0 ± 5.0, respectively, for patients without fracture, 54.8 ± 13.6 and 87.3 ± 10.8 for patients with a single fracture, and 53.1 ± 17.2 and 83.4 ± 12.7 for patients with multiple fractures. Patients with single and multiple fractures had similar (P = .91) preoperative IKDC scores but significantly lower (P < .001) postoperative IKDC scores compared to patients without fracture. Cortical depression fractures in patients with ACL tear are associated with decreased clinical outcome scores 1 year after ACL reconstruction surgery.