Aims: The aim of this study was to develop an ultra-short echo time 3D magnetic resonance imaging (MRI) method for imaging subacute myocardial infarction (MI) quantitatively and in an accelerated way. Here, we present novel 3D T1- and T1ρ-weighted Multi-Band SWeep Imaging with Fourier Transform and Compressed Sensing (MB-SWIFT-CS) imaging of subacute MI in mice hearts ex vivo. Methods and results: Relaxation time–weighted and under-sampled 3D MB-SWIFT-CS MRI were tested with manganese chloride (MnCl2) phantom and mice MI model. MI was induced in C57BL mice, and the hearts were collected 7 days after MI and then fixated. The hearts were imaged with T1 and adiabatic T1ρ relaxation time–weighted 3D MB-SWIFT-CS MRI, and the contrast-weighted image series were estimated with a locally low-rank regularized subspace constrained reconstruction. The quantitative parameter maps, T1 and T1ρ, were then obtained by performing non-linear least squares signal fitting on the image estimates. For comparison, the hearts were also imaged using 2D fast spin echo-based T2 and T1ρ mapping methods. The relaxation rates varied linearly with the MnCl2 concentration, and the T1 and T1ρ relaxation time values were elevated in the damaged areas. The ischaemic areas could be observed visually in the 3D T1, 3D T1ρ, and 2D MRI maps. The scar tissue formation in the anterior wall of the left ventricle and inflammation in the septum were confirmed by histology, which is in line with the results of MRI. Conclusion: MI with early fibrosis, increased inflammatory activity, and interstitial oedema were determined simultaneously with T1 and T1ρ relaxation time constants within the myocardium by using the 3D MB-SWIFT-CS method, allowing quantitative isotropic 3D assessment of the entire myocardium.

Knowledge of the relative performance of the well-known sparse and low-rank compressed sensing models with 3D radial quantitative magnetic resonance imaging acquisitions is limited. We use 3D radial T₁ relaxation time mapping data to compare the total variation, low-rank, and Huber penalty function approaches to regularization to provide insights into the relative performance of these image reconstruction models. Simulation and ex vivo specimen data were used to determine the best compressed sensing model as measured by normalized root mean squared error and structural similarity index. The large-scale compressed sensing models were solved by combining a GPU implementation of a preconditioned primal-dual proximal splitting algorithm to provide high-quality T₁ maps within a feasible computation time. The model combining spatial total variation and locally low-rank regularization yielded the best performance, followed closely by the model combining spatial and contrast dimension total variation. Computation times ranged from 2 to 113 min, with the low-rank approaches taking the most time. The differences between the compressed sensing models are not necessarily large, but the overall performance is heavily dependent on the imaged object.

Quantitative MRI (qMRI) methods allow reducing the subjectivity of clinical MRI by providing numerical values on which diagnostic assessment or predictions of tissue properties can be based. However, qMRI measurements typically take more time than anatomical imaging due to requiring multiple measurements with varying contrasts for, e.g., relaxation time mapping. To reduce the scanning time, undersampled data may be combined with compressed sensing (CS) reconstruction techniques. Typical CS reconstructions first reconstruct a complex-valued set of images corresponding to the varying contrasts, followed by a non-linear signal model fit to obtain the parameter maps. We propose a direct, embedded reconstruction method for T${}_{1\rho }$ mapping. The proposed method capitalizes on a known signal model to directly reconstruct the desired parameter map using a non-linear optimization model. The proposed reconstruction method also allows directly regularizing the parameter map of interest and greatly reduces the number of unknowns in the reconstruction, which are key factors in the performance of the reconstruction method. We test the proposed model using simulated radially sampled data from a 2D phantom and 2D cartesian ex vivo measurements of a mouse kidney specimen. We compare the embedded reconstruction model to two CS reconstruction models and in the cartesian test case also the direct inverse fast Fourier transform. The T${}_{1\rho }$ RMSE of the embedded reconstructions was reduced by 37–76% compared to the CS reconstructions when using undersampled simulated data with the reduction growing with larger acceleration factors. The proposed, embedded model outperformed the reference methods on the experimental test case as well, especially providing robustness with higher acceleration factors.

Purpose: To assess the potential for accelerating continuous‐wave (CW) T_1ρ dispersion measurement with compressed sensing approach via studying the effect that the data reduction has on the ability to detect differences between intact and degenerated articular cartilage with different spin‐lock amplitudes and to assess quantitative bias due to acceleration.Methods: Osteochondral plugs (n = 27, 4 mm diameter) from femur (n = 14) and tibia (n = 13) regions from human cadaver knee joints were obtained from commercial biobank (Science Care, USA) under Ethical permission 134/2015. MRI of specimens was performed at 9.4T with magnetization prepared radial balanced SSFP (bSSFP) readout sequence, and the CWT_1ρ relaxation time maps were computed from the measured data. The relaxation time maps were evaluated in the cartilage zones for different acceleration factors. For reference, Osteoarthritis Research Society International (OARSI) grading and biomechanical measurements were performed and correlated with the MRI findings.Results: Four‐fold acceleration of CWT_1ρ dispersion measurement by compressed sensing approach was feasible without meaningful loss in the sensitivity to osteoarthritic (OA) changes within the articular cartilage. Differences were significant between intact and OA groups in the superficial and transitional zones, and CWT_1ρ correlated moderately with the reference measurements (0.3 < r < 0.7).Conclusion: CWT_1ρ was able to differentiate between intact and OA cartilage even with four‐fold acceleration. This indicates that acceleration of CWT_1ρ dispersion measurement by compressed sensing approach is feasible with negligible loss in the sensitivity to osteoarthritic changes in articular cartilage.

Purpose: To determine how various compressed sensing (CS) models can accelerate alternating Look‐Locker mapping.Methods: An alternating Look‐Locker acquisition was retrospectively accelerated by factors of 1–12. The data was reconstructed into 12 images with multiple CS models, which utilized combinations of spatial total variation, locally low‐rank regularization, and subspace constraints. Complex non‐linear least squares signal fitting was performed to obtain the maps. The accelerated maps were compared against the map of a full data reference reconstruction.Results: A subspace‐constrained reconstruction model with spatial total variation and locally low‐rank regularization outperformed all other models as measured by map normalized root mean squared error, structural similarity index, and normalized mean absolute deviation. The subspace constraint benefited models utilizing spatial total variation but, conversely, did not benefit models utilizing only locally low‐rank regularization.Conclusion: The radial 3‐D alternating Look‐Locker mapping acquisition was successfully accelerated by up to a factor of 12 with various CS models. The best‐performing model was a subspace‐constrained reconstruction, which utilized spatial total variation and locally low‐rank regularization.