SPIRiT: Iterative self-consistent parallel imaging reconstruction from arbitrary k-space

Compressed sensing plus motion (CS + M): A new perspective for improving undersampled MR image reconstruction

We address the problem of reconstructing high quality images from undersampled MRI data. This is a challenging task due to the highly ill-posed nature of the problem. In particular, in dynamic MRI scans, the interaction between the target structure and the physical motion affects the acquired measurements leading to blurring artefacts and loss of fine details. In this work, we propose a framework for dynamic MRI reconstruction framed under a new multi-task optimisation model called Compressed Sensing Plus Motion (CS + M). Firstly, we propose a single optimisation problem that simultaneously computes the MRI reconstruction and the physical motion. Secondly, we show our model can be efficiently solved by breaking it up into two computationally tractable problems. The potentials and generalisation capabilities of our approach are demonstrated in different clinical applications including cardiac cine, cardiac perfusion and brain perfusion imaging. We show, through numerical experiments, that the proposed scheme reduces blurring artefacts, and preserves the target shape and fine details in the reconstruction. We also report the highest quality reconstruction under high undersampling rates in comparison to several state of the art techniques.

A Deep Framework Assembling Principled Modules for CS-MRI: Unrolling Perspective, Convergence Behaviors, and Practical Modeling

Compressed Sensing Magnetic Resonance Imaging (CS-MRI) significantly accelerates MR acquisition at a sampling rate much lower than the Nyquist criterion. A major challenge for CS-MRI lies in solving the severely ill-posed inverse problem to reconstruct aliasing-free MR images from the sparse k -space data. Conventional methods typically optimize an energy function, producing restoration of high quality, but their iterative numerical solvers unavoidably bring extremely large time consumption. Recent deep techniques provide fast restoration by either learning direct prediction to final reconstruction or plugging learned modules into the energy optimizer. Nevertheless, these data-driven predictors cannot guarantee the reconstruction following principled constraints underlying the domain knowledge so that the reliability of their reconstruction process is questionable. In this paper, we propose a deep framework assembling principled modules for CS-MRI that fuses learning strategy with the iterative solver of a conventional reconstruction energy. This framework embeds an optimal condition checking mechanism, fostering efficient and reliable reconstruction. We also apply the framework to three practical tasks, i.e., complex-valued data reconstruction, parallel imaging and reconstruction with Rician noise. Extensive experiments on both benchmark and manufacturer-testing images demonstrate that the proposed method reliably converges to the optimal solution more efficiently and accurately than the state-of-the-art in various scenarios.

Self-supervised learning of physics-guided reconstruction neural networks without fully sampled reference data

PURPOSE

To develop a strategy for training a physics-guided MRI reconstruction neural network without a database of fully sampled data sets.

METHODS

Self-supervised learning via data undersampling (SSDU) for physics-guided deep learning reconstruction partitions available measurements into two disjoint sets, one of which is used in the data consistency (DC) units in the unrolled network and the other is used to define the loss for training. The proposed training without fully sampled data is compared with fully supervised training with ground-truth data, as well as conventional compressed-sensing and parallel imaging methods using the publicly available fastMRI knee database. The same physics-guided neural network is used for both proposed SSDU and supervised training. The SSDU training is also applied to prospectively two-fold accelerated high-resolution brain data sets at different acceleration rates, and compared with parallel imaging.

RESULTS

Results on five different knee sequences at an acceleration rate of 4 shows that the proposed self-supervised approach performs closely with supervised learning, while significantly outperforming conventional compressed-sensing and parallel imaging, as characterized by quantitative metrics and a clinical reader study. The results on prospectively subsampled brain data sets, in which supervised learning cannot be used due to lack of ground-truth reference, show that the proposed self-supervised approach successfully performs reconstruction at high acceleration rates (4, 6, and 8). Image readings indicate improved visual reconstruction quality with the proposed approach compared with parallel imaging at acquisition acceleration.

CONCLUSION

The proposed SSDU approach allows training of physics-guided deep learning MRI reconstruction without fully sampled data, while achieving comparable results with supervised deep learning MRI trained on fully sampled data.

Super-resolution and denoising of 4D-Flow MRI using physics-Informed deep neural nets

BACKGROUND AND OBJECTIVE

Time resolved three-dimensional phase contrast magnetic resonance imaging (4D-Flow MRI) has been used to non-invasively measure blood velocities in the human vascular system. However, issues such as low spatio-temporal resolution, acquisition noise, velocity aliasing, and phase-offset artifacts have hampered its clinical application. In this research, we developed a purely data-driven method for super-resolution and denoising of 4D-Flow MRI.

METHODS

The flow velocities, pressure, and the MRI image magnitude are modeled as a patient-specific deep neural net (DNN). For training, 4D-Flow MRI images in the complex Cartesian space are used to impose data-fidelity. Physics of fluid flow is imposed through regularization. Creative loss function terms have been introduced to handle noise and super-resolution. The trained patient-specific DNN can be sampled to generate noise-free high-resolution flow images. The proposed method has been implemented using the TensorFlow DNN library and tested on numerical phantoms and validated in-vitro using high-resolution particle image velocitmetry (PIV) and 4D-Flow MRI experiments on transparent models subjected to pulsatile flow conditions.

RESULTS

In case of numerical phantoms, we were able to increase spatial resolution by a factor of 100 and temporal resolution by a factor of 5 compared to the simulated 4D-Flow MRI. There is an order of magnitude reduction of velocity normalized root mean square error (vNRMSE). In case of the in-vitro validation tests with PIV as reference, there is similar improvement in spatio-temporal resolution. Although the vNRMSE is reduced by 50%, the method is unable to negate a systematic bias with respect to the reference PIV that is introduced by the 4D-Flow MRI measurement.

CONCLUSIONS

This work has demonstrated the feasibility of using the readily available machinery of deep learning to enhance 4D-Flow MRI using a purely data-driven method. Unlike current state-of-the-art methods, the proposed method is agnostic to geometry and boundary conditions and therefore eliminates the need for tedious tasks such as accurate image segmentation for geometry, image registration, and estimation of boundary flow conditions. Arbitrary regions of interest can be selected for processing. This work will lead to user-friendly analysis tools that will enable quantitative hemodynamic analysis of vascular diseases in a clinical setting.

A dictionary-based graph-cut algorithm for MRI reconstruction

PURPOSE

Compressive sensing (CS)-based image reconstruction methods have proposed random undersampling schemes that produce incoherent, noise-like aliasing artifacts, which are easier to remove. The denoising process is critically assisted by imposing sparsity-enforcing priors. Sparsity is known to be induced if the prior is in the form of the L_p (0 ≤ p ≤ 1) norm. CS methods generally use a convex relaxation of these priors such as the L₁ norm, which may not exploit the full power of CS. An efficient, discrete optimization formulation is proposed, which works not only on arbitrary L_p -norm priors as some non-convex CS methods do, but also on highly non-convex truncated penalty functions, resulting in a specific type of edge-preserving prior. These advanced features make the minimization problem highly non-convex, and thus call for more sophisticated minimization routines.

THEORY AND METHODS

The work combines edge-preserving priors with random undersampling, and solves the resulting optimization using a set of discrete optimization methods called graph cuts. The resulting optimization problem is solved by applying graph cuts iteratively within a dictionary, defined here as an appropriately constructed set of vectors relevant to brain MRI data used here.

RESULTS

Experimental results with in vivo data are presented.

CONCLUSION

The proposed algorithm produces better results than regularized SENSE or standard CS for reconstruction of in vivo data.

Parallel imaging with a combination of sensitivity encoding and generative adversarial networks

Background

Magnetic resonance imaging (MRI) has the limitation of low imaging speed. Acceleration methods using under-sampled k-space data have been widely exploited to improve data acquisition without reducing the image quality. Sensitivity encoding (SENSE) is the most commonly used method for multi-channel imaging. However, SENSE has the drawback of severe g-factor artifacts when the under-sampling factor is high. This paper applies generative adversarial networks (GAN) to remove g-factor artifacts from SENSE reconstructions.

Methods

Our method was evaluated on a public knee database containing 20 healthy participants. We compared our method with conventional GAN using zero-filled (ZF) images as input. Structural similarity (SSIM), peak signal to noise ratio (PSNR), and normalized mean square error (NMSE) were calculated for the assessment of image quality. A paired student's t-test was conducted to compare the image quality metrics between the different methods. Statistical significance was considered at P<0.01.

Results

The proposed method outperformed SENSE, variational network (VN), and ZF + GAN methods in terms of SSIM (SENSE + GAN: 0.81±0.06, SENSE: 0.40±0.07, VN: 0.79±0.06, ZF + GAN: 0.77±0.06), PSNR (SENSE + GAN: 31.90±1.66, SENSE: 22.70±1.99, VN: 31.35±2.01, ZF + GAN: 29.95±1.59), and NMSE (×10^-7) (SENSE + GAN: 0.95±0.34, SENSE: 4.81±1.33, VN: 0.97±0.30, ZF + GAN: 1.60±0.84) with an under-sampling factor of up to 6-fold.

Conclusions

This study demonstrated the feasibility of using GAN to improve the performance of SENSE reconstruction. The improvement of reconstruction is more obvious for higher under-sampling rates, which shows great potential for many clinical applications.

Deep Generalization of Structured Low-Rank Algorithms (Deep-SLR)

Structured low-rank (SLR) algorithms, which exploit annihilation relations between the Fourier samples of a signal resulting from different properties, is a powerful image reconstruction framework in several applications. This scheme relies on low-rank matrix completion to estimate the annihilation relations from the measurements. The main challenge with this strategy is the high computational complexity of matrix completion. We introduce a deep learning (DL) approach to significantly reduce the computational complexity. Specifically, we use a convolutional neural network (CNN)-based filterbank that is trained to estimate the annihilation relations from imperfect (under-sampled and noisy) k-space measurements of Magnetic Resonance Imaging (MRI). The main reason for the computational efficiency is the pre-learning of the parameters of the non-linear CNN from exemplar data, compared to SLR schemes that learn the linear filterbank parameters from the dataset itself. Experimental comparisons show that the proposed scheme can enable calibration-less parallel MRI; it can offer performance similar to SLR schemes while reducing the runtime by around three orders of magnitude. Unlike pre-calibrated and self-calibrated approaches, the proposed uncalibrated approach is insensitive to motion errors and affords higher acceleration. The proposed scheme also incorporates image domain priors that are complementary, thus significantly improving the performance over that of SLR schemes.

Highly accelerated submillimeter resolution 3D GRASE with controlled T 2 blurring in T 2 -weighted functional MRI at 7 Tesla: A feasibility study

PURPOSE

To achieve highly accelerated submillimeter resolution T 2 -weighted functional MRI at 7T by developing a three-dimensional gradient and spin echo imaging (GRASE) with inner-volume selection and variable flip angles (VFA).

METHODS

GRASE imaging has disadvantages in that (a) k-space modulation causes T 2 blurring by limiting the number of slices and (b) a VFA scheme results in partial success with substantial SNR loss. In this work, accelerated GRASE with controlled T 2 blurring is developed to improve a point spread function (PSF) and temporal signal-to-noise ratio (tSNR) with a large number of slices. To this end, the VFA scheme is designed by minimizing a trade-off between SNR and blurring for functional sensitivity, and a new GRASE-optimized random encoding, which takes into account the complex signal decays of T 2 and T 2 ∗ weightings, is proposed by achieving incoherent aliasing for constrained reconstruction. Numerical and experimental studies were performed to validate the effectiveness of the proposed method over regular and VFA GRASE (R- and V-GRASE).

RESULTS

The proposed method, while achieving 0.8 mm isotropic resolution, functional MRI compared to R- and V-GRASE improves the spatial extent of the excited volume up to 36 slices with 52%-68% full width at half maximum (FWHM) reduction in PSF but approximately 2- to 3-fold mean tSNR improvement, thus resulting in higher BOLD activations.

CONCLUSIONS

We successfully demonstrated the feasibility of the proposed method in T 2 -weighted functional MRI. The proposed method is especially promising for cortical layer-specific functional MRI.

Real-time exercise stress cardiac MRI with Fourier-series reconstruction from golden-angle radial data

Magnetic resonance imaging (MRI) can measure cardiac response to exercise stress for evaluating and managing heart patients in the practice of clinical cardiology. However, exercise stress cardiac MRI have been clinically limited by the ability of available MRI techniques to quantitatively measure fast and unstable cardiac dynamics during exercise. The presented work is to develop a new real-time MRI technique for improved quantitative performance of exercise stress cardiac MRI. This technique seeks to represent real-time cardiac images as a sparse Fourier-series along the time. With golden-angle radial acquisition, parallel imaging and compressed sensing can be integrated into a linear system of equations for resolving Fourier coefficients that are in turn used to generate real-time cardiac images from the Fourier-series representation. Fourier-series reconstruction from golden-angle radial data can effectively address data insufficiency due to MRI speed limitation, providing a real-time approach to exercise stress cardiac MRI. To demonstrate the feasibility, an exercise stress cardiac MRI experiment was run to investigate biventricular response to in-scanner biking exercise in a cohort of sixteen healthy volunteers. It was found that Fourier-series reconstruction from golden-angle radial data effectively detected exercise-induced increase in stroke volume and ejection fraction in a healthy heart. The presented work will improve the applications of exercise stress cardiac MRI in the practice of clinical cardiology.

Multiple Slice k-space Deep Learning for Magnetic Resonance Imaging Reconstruction

Magnetic resonance imaging (MRI) has been one of the most powerful and valuable imaging methods for medical diagnosis and staging of disease. Due to the long scan time of MRI acquisition, k-space under-samplings is required during the acquisition processing. Thus, MRI reconstruction, which transfers undersampled k-space data to high-quality magnetic resonance imaging, becomes an important and meaningful task. There have been many explorations on k-space interpolation for MRI reconstruction. However, most of these methods ignore the strong correlation between target slice and its adjacent slices. Inspired by this, we propose a fully data-driven deep learning algorithm for k-space interpolation, utilizing the correlation information between the target slice and its neighboring slices. A novel network is proposed, which models the inter-dependencies between different slices. In addition, the network is easily implemented and expended. Experiments show that our methods consistently surpass existing image-domain and k-space-domain magnetic resonance imaging reconstructing methods.

Three-dimensional Yarnball k-space acquisition for accelerated MRI

PURPOSE

To introduce an efficient sampling technique named Yarnball, which may serve as a direct alternative to 3D Cones.

METHODS

Yarnball evolves through 3D k-space with increasing loop size, and the differential equations defining this flexible trajectory are presented in detail. The sampling efficiencies of Yarnball and 3D Cones were compared through point spread function analysis and simulated imaging (which highlights undersampling in the absence of other scanning effects). The feasibility of Yarnball implementation was demonstrated for fully sampled T₁ -weighted images of the human head at 3 T.

RESULTS

The mostly large 3D loops of the Yarnball trajectory facilitate rapid sampling under peripheral nerve stimulation constraint, an advantage that increases with readout duration (T_RO ). Point spread function analysis yielded 89% (T_RO = 2 ms) and 77% (T_RO = 10 ms) of Yarnball voxels with magnitude less than 0.01% of the point spread function peak. For 3D Cones, these values were only 52% and 29%. The 3D-Cones technique required 1.4 times (T_RO = 2 ms) and 1.8 times (T_RO = 10 ms) more trajectories than Yarnball to produce simulated images of a sphere free from undersampling artifact. For a prolate spheroidal (head-like) object, 1.75 times and 2.6 times more trajectories were required for 3D Cones. Yarnball produced 0.72 mm (1/2k_max ) isotropic T₁ -weighted human brain images free from undersampling artifact in only 98 seconds at 3 T.

CONCLUSION

Yarnball demonstrated greater k-space sampling efficiency than directly comparable 3D Cones, and may have value wherever 3D Cones has been considered. Yarnball may also have value in the context of rapid T₁ -weighted brain imaging.

J-MoDL: Joint Model-Based Deep Learning for Optimized Sampling and Reconstruction

Modern MRI schemes, which rely on compressed sensing or deep learning algorithms to recover MRI data from undersampled multichannel Fourier measurements, are widely used to reduce the scan time. The image quality of these approaches is heavily dependent on the sampling pattern. We introduce a continuous strategy to optimize the sampling pattern and the network parameters jointly. We use a multichannel forward model, consisting of a non-uniform Fourier transform with continuously defined sampling locations, to realize the data consistency block within a model-based deep learning image reconstruction scheme. This approach facilitates the joint and continuous optimization of the sampling pattern and the CNN parameters to improve image quality. We observe that the joint optimization of the sampling patterns and the reconstruction module significantly improves the performance of most deep learning reconstruction algorithms. The source code is available at https://github.com/hkaggarwal/J-MoDL.

Non-contrast coronary magnetic resonance angiography: current frontiers and future horizons

Coronary magnetic resonance angiography (coronary MRA) is advantageous in its ability to assess coronary artery morphology and function without ionizing radiation or contrast media. However, technical limitations including reduced spatial resolution, long acquisition times, and low signal-to-noise ratios prevent it from clinical routine utilization. Nonetheless, each of these limitations can be specifically addressed by a combination of novel technologies including super-resolution imaging, compressed sensing, and deep-learning reconstruction. In this paper, we first review the current clinical use and motivations for non-contrast coronary MRA, discuss currently available coronary MRA techniques, and highlight current technical developments that hold unique potential to optimize coronary MRA image acquisition and post-processing. In the final section, we examine the various research-based coronary MRA methods and metrics that can be leveraged to assess coronary stenosis severity, physiological function, and atherosclerotic plaque characterization. We specifically discuss how such technologies may contribute to the clinical translation of coronary MRA into a robust modality for routine clinical use.

Calibrationless parallel imaging reconstruction for multislice MR data using low-rank tensor completion

PURPOSE

To provide joint calibrationless parallel imaging reconstruction of highly accelerated multislice 2D MR k-space data.

METHODS

Adjacent image slices in multislice MR data have similar coil sensitivity maps, spatial support, and image content. Such similarities can be utilized to improve image quality by reconstructing multiple slices jointly with low-rank tensor completion. Specifically, the multichannel k-space data from multiple slices are constructed into a block-wise Hankel tensor and iteratively updated by promoting tensor low-rankness through higher-order SVD. This multislice block-wise Hankel tensor completion was implemented for 2D spiral and Cartesian k-space undersampling where sampling patterns vary between adjacent slices. The approach was evaluated with human brain MR data and compared to the traditional single-slice simultaneous autocalibrating and k-space estimation reconstruction.

RESULTS

The proposed multislice block-wise Hankel tensor completion approach robustly reconstructed highly undersampled multislice 2D spiral and Cartesian data. It produced substantially lower level of artifacts compared to the traditional single-slice simultaneous autocalibrating and k-space estimation reconstruction. Quantitative evaluation using error maps and root mean square error demonstrated its significantly improved performance in terms of residual artifacts and root mean square error.

CONCLUSION

Our proposed multislice block-wise Hankel tensor completion method exploits the similar coil sensitivity and image content within multislice MR data through a tensor completion framework. It offers a new and effective approach to acquire and reconstruct highly undersampled multislice MR data in a calibrationless manner.

Detecting small pulmonary nodules with spiral ultrashort echo time sequences in 1.5 T MRI

OBJECTIVE

This study investigated ultrashort echo time (UTE) sequences in 1.5 T magnetic resonance imaging (MRI) for small lung nodule detection.

MATERIALS AND METHODS

A total of 120 patients with 165 small lung nodules before video-associated thoracoscopic resection were enrolled. MRI sequences included conventional volumetric interpolated breath-hold examination (VIBE, scan time 16 s), spiral UTE (TE 0.05 ms) with free-breathing (scan time 3.5-5 min), and breath-hold sequences (scan time 20 s). Chest CT provided a standard reference for nodule size and morphology. Nodule detection sensitivity was evaluated on a lobe-by-lobe basis.

RESULTS

The nodule detection rate was significantly higher in spiral UTE free-breathing (> 78%, p < 0.05) and breath-hold sequences (> 75%, p < 0.05) compared with conventional VIBE (> 55%), reaching 100% when nodule size was > 16 mm, and reaching 95% when nodules were in solid morphology, regardless of size. The inter-sequence reliability between free-breathing and breath-hold spiral UTE was good (κ > 0.80). Inter-reader agreement was also high (κ > 0.77) for spiral UTE sequences. Nodule size measurements were consistent between CT and spiral UTE MRI, with a minimal bias up to 0.2 mm.

DISCUSSION

Spiral UTE sequences detect small lung nodules that warrant surgery, offers realistic scan times for clinical work, and could be implemented as part of routine lung MRI.

Diagnostic quality assessment of IR-prepared 3D magnetic resonance neuroimaging accelerated using compressed sensing and k-space sampling order optimization

PURPOSE

To evaluate the clinical diagnostic efficacy of accelerated 3D magnetic resonance (MR) neuroimaging by radiological assessment for image quality and artefacts.

STUDY TYPE

Prospective healthy volunteer study.

SUBJECTS

Eight healthy subjects.

FIELD STRENGTH/SEQUENCE

Inversion Recovery (IR) prepared 3D Gradient Echo (GRE) sequence on a 1.5 T GE Signa HDx scanner.

ASSESSMENT

Independent radiological diagnostic quality assessments of accelerated 3D MR brain datasets were carried out by four experienced neuro-radiologists who were blinded to the acceleration factor and to the subject. The radiological grading was based on a previously reported radiological scoring key that was used for image quality assessment of human brains.

STATISTICAL TESTS

Bland-Altman analysis.

RESULTS

Optimization of the k-space sampling order was important for preserving contrast in accelerated scans. Despite having lower scores than fully sampled datasets, the majority of the compressed sensing (CS) accelerated brain datasets with k-space sampling order optimization (19/24 datasets by Radiologist 1, 24/24 datasets by Radiologist 2 and 16/24 datasets by Radiologist 3) were graded to be fully diagnostic indicating that there was adequate confidence for performing gross structural assessment of the brain.

CONCLUSION

Optimization of k-space acquisition order improves the clinical utility of CS accelerated 3D neuroimaging. This method may be appropriate for routine radiological assessment of the brain.

Improving Quantitative Magnetic Resonance Imaging Using Deep Learning

Deep learning methods have shown promising results for accelerating quantitative musculoskeletal (MSK) magnetic resonance imaging (MRI) for T2 and T1ρ relaxometry. These methods have been shown to improve musculoskeletal tissue segmentation on parametric maps, allowing efficient and accurate T2 and T1ρ relaxometry analysis for monitoring and predicting MSK diseases. Deep learning methods have shown promising results for disease detection on quantitative MRI with diagnostic performance superior to conventional machine-learning methods for identifying knee osteoarthritis.

SURE-based automatic parameter selection for ESPIRiT calibration

PURPOSE

ESPIRiT is a parallel imaging method that estimates coil sensitivity maps from the auto-calibration region (ACS). This requires choosing several parameters for the optimal map estimation. While fairly robust to these parameter choices, occasionally, poor selection can result in reduced performance. The purpose of this work is to automatically select parameters in ESPIRiT for more robust and consistent performance across a variety of exams.

METHODS

By viewing ESPIRiT as a denoiser, Stein's unbiased risk estimate (SURE) is leveraged to automatically optimize parameter selection in a data-driven manner. The optimum parameters corresponding to the minimum true squared error, minimum SURE as derived from densely sampled, high-resolution, and non-accelerated data and minimum SURE as derived from ACS are compared using simulation experiments. To avoid optimizing the rank of ESPIRiT's auto-calibrating matrix (one of the parameters), a heuristic derived from SURE-based singular value thresholding is also proposed.

RESULTS

Simulations show SURE derived from the densely sampled, high-resolution, and non-accelerated data to be an accurate estimator of the true mean squared error, enabling automatic parameter selection. The parameters that minimize SURE as derived from ACS correspond well to the optimal parameters. The soft-threshold heuristic improves computational efficiency while providing similar results to an exhaustive search. In-vivo experiments verify the reliability of this method.

CONCLUSIONS

Using SURE to determine ESPIRiT parameters allows for automatic parameter selections. In-vivo results are consistent with simulation and theoretical results.

Deep complex convolutional network for fast reconstruction of 3D late gadolinium enhancement cardiac MRI

Several deep-learning models have been proposed to shorten MRI scan time. Prior deep-learning models that utilize real-valued kernels have limited capability to learn rich representations of complex MRI data. In this work, we utilize a complex-valued convolutional network (ℂNet) for fast reconstruction of highly under-sampled MRI data and evaluate its ability to rapidly reconstruct 3D late gadolinium enhancement (LGE) data. ℂNet preserves the complex nature and optimal combination of real and imaginary components of MRI data throughout the reconstruction process by utilizing complex-valued convolution, novel radial batch normalization, and complex activation function layers in a U-Net architecture. A prospectively under-sampled 3D LGE cardiac MRI dataset of 219 patients (17 003 images) at acceleration rates R = 3 through R = 5 was used to evaluate ℂNet. The dataset was further retrospectively under-sampled to a maximum of R = 8 to simulate higher acceleration rates. We created three reconstructions of the 3D LGE dataset using (1) ℂNet, (2) a compressed-sensing-based low-dimensional-structure self-learning and thresholding algorithm (LOST), and (3) a real-valued U-Net (realNet) with the same number of parameters as ℂNet. LOST-reconstructed data were considered the reference for training and evaluation of all models. The reconstructed images were quantitatively evaluated using mean-squared error (MSE) and the structural similarity index measure (SSIM), and subjectively evaluated by three independent readers. Quantitatively, ℂNet-reconstructed images had significantly improved MSE and SSIM values compared with realNet (MSE, 0.077 versus 0.091; SSIM, 0.876 versus 0.733, respectively; p < 0.01). Subjective quality assessment showed that ℂNet-reconstructed image quality was similar to that of compressed sensing and significantly better than that of realNet. ℂNet reconstruction was also more than 300 times faster than compressed sensing. Retrospective under-sampled images demonstrate the potential of ℂNet at higher acceleration rates. ℂNet enables fast reconstruction of highly accelerated 3D MRI with superior performance to real-valued networks, and achieves faster reconstruction than compressed sensing.

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

Long scan duration remains a challenge for high-resolution MRI. Deep learning has emerged as a powerful means for accelerated MRI reconstruction by providing data-driven regularizers that are directly learned from data. These data-driven priors typically remain unchanged for future data in the testing phase once they are learned during training. In this study, we propose to use a transfer learning approach to fine-tune these regularizers for new subjects using a self-supervision approach. While the proposed approach can compromise the extremely fast reconstruction time of deep learning MRI methods, our results on knee MRI indicate that such adaptation can substantially reduce the remaining artifacts in reconstructed images. In addition, the proposed approach has the potential to reduce the risks of generalization to rare pathological conditions, which may be unavailable in the training data.

Temperature estimation for MR-guided microwave hyperthermia using block-based compressed sensing. ANNUAL INTERNATIONAL CONFERENCE OF THE IEEE ENGINEERING IN MEDICINE AND BIOLOGY SOCIETY. IEEE ENGINEERING IN MEDICINE AND BIOLOGY SOCIETY. ANNUAL INTERNATIONAL CONFERENCE 2020;2020:5057-5060. [PMID: 33019123 DOI: 10.1109/embc44109.2020.9176206] [Citation(s) in RCA: 0] [Impact Index Per Article: 0]

Mild hyperthermia has been clinically employed as an adjuvant for radiation/chemotherapy and is under investigation for precise thermally-mediated delivery of cancer therapeutic agents. Magnetic Resonance Imaging (MRI) facilitates non-invasive, real-time spatial thermometry for monitoring and guiding hyperthermia procedures. Long image acquisition time during MR-guided hyperthermia may fail to capture rapid changes in temperature. This may lead to unwanted heating of healthy tissue and/or temperature rise above hyperthermic range. We have developed a block-based compressed sensing approach to reconstruct volumetric MR-derived microwave hyperthermia temperature profiles using a subset of measured data. This algorithm exploits the sparsity of MR images due to the presence of inter- and intra-slice correlation of hyperthermic MR-derived temperature profiles. We have evaluated the performance of our developed algorithm on a phantom and in vivo in mice using previously implemented microwave applicators. This algorithm reconstructs 3D temperature profiles with PSNR of 33 dB - 49 dB in comparison to the original profiles. In summary, this study suggests that microwave hyperthermia induced temperature profiles can be reconstructed using subsamples to reduce MR image acquisition time.

Navigator-based reacquisition and estimation of motion-corrupted data: Application to multi-echo spin echo for carotid wall MRI

PURPOSE

To assess whether artifacts in multi-slice multi-echo spin echo neck imaging, thought to be caused by brief motion events such as swallowing, can be corrected by reacquiring corrupted central k-space data and estimating the remainder with parallel imaging.

METHODS

A single phase-encode line (ky = 0, phase-encode direction anteroposterior) navigator echo was used to identify motion-corrupted data and guide the online reacquisition. If motion corruption was detected in the 7 central k-space lines, they were replaced with reacquired data. Subsequently, GRAPPA reconstruction was trained on the updated central portion of k-space and then used to estimate the remaining motion-corrupted k-space data from surrounding uncorrupted data. Similar compressed sensing-based approaches have been used previously to compensate for respiration in cardiac imaging. The g-factor noise amplification was calculated for the parallel imaging reconstruction of data acquired with a 10-channel neck coil. The method was assessed in scans with 9 volunteers and 12 patients.

RESULTS

The g-factor analysis showed that GRAPPA reconstruction of 2 adjacent motion-corrupted lines causes high noise amplification; therefore, the number of 2-line estimations should be limited. In volunteer scans, median ghosting reduction of 24% was achieved with 2 adjacent motion-corrupted lines correction, and image quality was improved in 2 patient scans that had motion corruption close to the center of k-space.

CONCLUSION

Motion-corrupted echo-trains can be identified with a navigator echo. Combined reacquisition and parallel imaging estimation reduced motion artifacts in multi-slice MESE when there were brief motion events, especially when motion corruption was close to the center of k-space.

Three-dimensional Ultrashort Echotime Magnetic Resonance Imaging for Combined Morphologic and Ventilation Imaging in Pediatric Patients With Pulmonary Disease

PURPOSE

Ultrashort echotime (UTE) sequences aim to improve the signal yield in pulmonary magnetic resonance imaging (MRI). We demonstrate the initial results of spiral 3-dimensional (3D) UTE-MRI for combined morphologic and functional imaging in pediatric patients.

METHODS

Seven pediatric patients with pulmonary abnormalities were included in this observational, prospective, single-center study, with the patients having the following conditions: cystic fibrosis (CF) with middle lobe atelectasis, CF with allergic bronchopulmonary aspergillosis, primary ciliary dyskinesia, air trapping, congenital lobar overinflation, congenital pulmonary airway malformation, and pulmonary hamartoma.Patients were scanned during breath-hold in 5 breathing states on a 3-Tesla system using a prototypical 3D stack-of-spirals UTE sequence. Ventilation maps and signal intensity maps were calculated. Morphologic images, ventilation-weighted maps, and signal intensity maps of the lungs of each patient were assessed intraindividually and compared with reference examinations.

RESULTS

With a scan time of ∼15 seconds per breathing state, 3D UTE-MRI allowed for sufficient imaging of both "plus" pathologies (atelectasis, inflammatory consolidation, and pulmonary hamartoma) and "minus" pathologies (congenital lobar overinflation, congenital pulmonary airway malformation, and air trapping). Color-coded maps of normalized signal intensity and ventilation increased diagnostic confidence, particularly with regard to "minus" pathologies. UTE-MRI detected new atelectasis in an asymptomatic CF patient, allowing for rapid and successful therapy initiation, and it was able to reproduce atelectasis and hamartoma known from multidetector computed tomography and to monitor a patient with allergic bronchopulmonary aspergillosis.

CONCLUSION

3D UTE-MRI using a stack-of-spirals trajectory enables combined morphologic and functional imaging of the lungs within ~115 second acquisition time and might be suitable for monitoring a wide spectrum of pulmonary diseases.

Accelerating Non-Cartesian MRI Reconstruction Convergence Using k-Space Preconditioning

We propose a k-space preconditioning formulation for accelerating the convergence of iterative Magnetic Resonance Imaging (MRI) reconstructions from non-uniformly sampled k-space data. Existing methods either use sampling density compensations which sacrifice reconstruction accuracy, or circulant preconditioners which increase per-iteration computation. Our approach overcomes both shortcomings. Concretely, we show that viewing the reconstruction problem in the dual formulation allows us to precondition in k-space using density-compensation-like operations. Using the primal-dual hybrid gradient method, the proposed preconditioning method does not have inner loops and are competitive in accelerating convergence compared to existing algorithms. We derive l2 -optimized preconditioners, and demonstrate through experiments that the proposed method converges in about ten iterations in practice.

Non-rigid image registration of 4D-MRI data for improved delineation of moving tumors

BACKGROUND

To increase the image quality of end-expiratory and end-inspiratory phases of retrospective respiratory self-gated 4D MRI data sets using non-rigid image registration for improved target delineation of moving tumors.

METHODS

End-expiratory and end-inspiratory phases of volunteer and patient 4D MRI data sets are used as targets for non-rigid image registration of all other phases using two different registration schemes: In the first, all phases are registered directly (dir-Reg) while next neighbors are successively registered until the target is reached in the second (nn-Reg). Resulting data sets are quantitatively compared using diaphragm and tumor sharpness and the coefficient of variation of regions of interest in the lung, liver, and heart. Qualitative assessment of the patient data regarding noise level, tumor delineation, and overall image quality was performed by blinded reading based on a 4 point Likert scale.

RESULTS

The median coefficient of variation was lower for both registration schemes compared to the target. Median dir-Reg coefficient of variation of all ROIs was 5.6% lower for expiration and 7.0% lower for inspiration compared with nn-Reg. Statistical significant differences between the two schemes were found in all comparisons. Median sharpness in inspiration is lower compared to expiration sharpness in all cases. Registered data sets were rated better compared to the targets in all categories. Over all categories, mean expiration scores were 2.92 ± 0.18 for the target, 3.19 ± 0.22 for nn-Reg and 3.56 ± 0.14 for dir-Reg and mean inspiration scores 2.25 ± 0.12 for the target, 2.72 ± 215 0.04 for nn-Reg and 3.78 ± 0.04 for dir-Reg.

CONCLUSIONS

In this work, end-expiratory and inspiratory phases of a 4D MRI data sets are used as targets for non-rigid image registration of all other phases. It is qualitatively and quantitatively shown that image quality of the targets can be significantly enhanced leading to improved target delineation of moving tumors.

Factorized sensitivity estimation for artifact suppression in phase-cycled bSSFP MRI

OBJECTIVE

Balanced steady-state free precession (bSSFP) imaging suffers from banding artifacts in the presence of magnetic field inhomogeneity. The purpose of this study is to identify an efficient strategy to reconstruct banding-free bSSFP images from multi-coil multi-acquisition datasets.

METHOD

Previous techniques either assume that a naïve coil-combination is performed a priori resulting in suboptimal artifact suppression, or that artifact suppression is performed for each coil separately at the expense of significant computational burden. Here we propose a tailored method that factorizes the estimation of coil and bSSFP sensitivity profiles for improved accuracy and/or speed.

RESULTS

In vivo experiments show that the proposed method outperforms naïve coil-combination and coil-by-coil processing in terms of both reconstruction quality and time.

CONCLUSION

The proposed method enables computationally efficient artifact suppression for phase-cycled bSSFP imaging with modern coil arrays. Rapid imaging applications can efficiently benefit from the improved robustness of bSSFP imaging against field inhomogeneity.

CONVOLUTIONAL FRAMEWORK FOR ACCELERATED MAGNETIC RESONANCE IMAGING

Magnetic Resonance Imaging (MRI) is a noninvasive imaging technique that provides exquisite soft-tissue contrast without using ionizing radiation. The clinical application of MRI may be limited by long data acquisition times; therefore, MR image reconstruction from highly undersampled k-space data has been an active area of research. Many works exploit rank deficiency in a Hankel data matrix to recover unobserved k-space samples; the resulting problem is non-convex, so the choice of numerical algorithm can significantly affect performance, computation, and memory. We present a simple, scalable approach called Convolutional Framework (CF). We demonstrate the feasibility and versatility of CF using measured data from 2D, 3D, and dynamic applications.

CALIBRATIONLESS PARALLEL MRI USING MODEL BASED DEEP LEARNING (C-MODL)

We introduce a fast model based deep learning approach for calibrationless parallel MRI reconstruction. The proposed scheme is a non-linear generalization of structured low rank (SLR) methods that self learn linear annihilation filters from the same subject. It pre-learns non-linear annihilation relations in the Fourier domain from exemplar data. The pre-learning strategy significantly reduces the computational complexity, making the proposed scheme three orders of magnitude faster than SLR schemes. The proposed framework also allows the use of a complementary spatial domain prior; the hybrid regularization scheme offers improved performance over calibrated image domain MoDL approach. The calibrationless strategy minimizes potential mismatches between calibration data and the main scan, while eliminating the need for a fully sampled calibration region.

Feasibility of 4D T2* quantification in the lung with oxygen gas challenge in patients with non-small cell lung cancer

Blood oxygen level-dependent (BOLD) MRI is a non-invasive diagnostic method for assessing tissue oxygenation level, by changes in the transverse relaxation time T2*. 3D BOLD imaging of lung tumours is challenging, because respiratory motion can lead to significant image quality degradation. The purpose of this work was to explore the feasibility of a three dimensional (3D) Cartesian multi gradient echo (MGRE) sequence for T2* measurements of non-small cell lung tumours during free-breathing. A non-uniform quasi-random reordering of the pahse encoding lines that allocates more sampling points near the k-space origin resulting in efficient undersampling pattern for parallel imaging was combined with multi echo acquisition and self-gating. In a series of three patients 3D T2* maps of lung carcinomas were generated with isotropic spatial resolution and full tumour coverage at air inhalation and after hyperoxic gas challenge in arbitrary respiratory phases using the proposed self-gated MGRE acquisition. The changes in T2* on the inhalation of hyperoxic gas relative to air were quantified. Significant changes in T2* were observed following oxygen inhalation in the tumour (p < 0.02). Thus, the self-gated MGRE sequence can be used for assessment of BOLD signal with isotropic resolution and arbitrary respiratory phases in non-small cell lung cancer.

Accelerated coronary MRI with sRAKI: A database-free self-consistent neural network k-space reconstruction for arbitrary undersampling

Purpose

To accelerate coronary MRI acquisitions with arbitrary undersampling patterns by using a novel reconstruction algorithm that applies coil self-consistency using subject-specific neural networks.

Methods

Self-consistent robust artificial-neural-networks for k-space interpolation (sRAKI) performs iterative parallel imaging reconstruction by enforcing self-consistency among coils. The approach bears similarity to SPIRiT, but extends the linear convolutions in SPIRiT to nonlinear interpolation using convolutional neural networks (CNNs). These CNNs are trained individually for each scan using the scan-specific autocalibrating signal (ACS) data. Reconstruction is performed by imposing the learned self-consistency and data-consistency, which enables sRAKI to support random undersampling patterns. Fully-sampled targeted right coronary artery MRI was acquired in six healthy subjects. The data were retrospectively undersampled, and reconstructed using SPIRiT, l₁-SPIRiT and sRAKI for acceleration rates of 2 to 5. Additionally, prospectively undersampled whole-heart coronary MRI was acquired to further evaluate reconstruction performance.

Results

sRAKI reduces noise amplification and blurring artifacts compared with SPIRiT and l₁-SPIRiT, especially at high acceleration rates in targeted coronary MRI. Quantitative analysis shows that sRAKI outperforms these techniques in terms of normalized mean-squared-error (~44% and ~21% over SPIRiT and l1-SPIRiT at rate 5) and vessel sharpness (~10% and ~20% over SPIRiT and l₁-SPIRiT at rate 5). Whole-heart data shows the sharpest coronary arteries when resolved using sRAKI, with 11% and 15% improvement in vessel sharpness over SPIRiT and l₁-SPIRiT, respectively.

Conclusion

sRAKI is a database-free neural network-based reconstruction technique that may further accelerate coronary MRI with arbitrary undersampling patterns, while improving noise resilience over linear parallel imaging and image sharpness over l₁ regularization techniques.

A locally segmented reconstruction method for parallel imaging

PURPOSE

A locally segmented parallel imaging reconstruction method is proposed that efficiently utilizes sensitivity distribution of multichannel receiver coil.

THEORY AND METHODS

A method of locally segmenting a MR signal is introduced to maximize the differences in sensitivity between receiver channels. A 1D Fourier transformation of the undersampled k-space data is performed along the readout direction, which generates a hybrid 2D space. The hybrid space is partitioned into localized segments along the readout direction. In every localized segment, kernels representing relation between adjacent signals are estimated from autocalibration signals, and data at unsampled points are estimated using the kernels. Then, the images are reconstructed from full k-space data that consists of the sampled data and the estimated data at unsampled points.

RESULTS

In a computer simulation and in vivo experiments, the locally segmented reconstruction method produced fewer residual artifacts compared to the conventional parallel imaging reconstruction methods with the same kernel geometry. The performance gain of the proposed method comes from maximizing encoding capability of receiver channels, thus resulting in the accurately estimated kernel weights that reflect the relation between adjacent signals.

CONCLUSION

The proposed spatial segmentation method maximally utilizes differences in the sensitivity of receiver channels to reconstruct images with reduced artifacts.

Time-domain principal component reconstruction (tPCR): A more efficient and stable iterative reconstruction framework for non-Cartesian functional MRI

PURPOSE

To improve the reconstruction efficiency (i.e., computational load) and stability of iterative reconstruction for non-Cartesian fMRI when using high undersampling rates and/or in the presence of strong off-resonance effects.

THEORY AND METHODS

The magnetic resonance encephalography (MREG) sequence with 3D non-Cartesian trajectory and 0.1s repetition time (TR) was applied to acquire fMRI datasets. Different from a conventional time-point-by-time-point sequential reconstruction (SR), the proposed time-domain principal component reconstruction (tPCR) performs three steps: (1) decomposing the k-t-space fMRI datasets into time-domain principal component space using singular value decomposition, (2) reconstructing each principal component with redistributed computation power according to their weights, and (3) combining the reconstructed principal components back to image-t-space. The comparison of reconstruction accuracy was performed by simulation experiments and then verified in real fMRI data.

RESULTS

The simulation experiments showed that the proposed tPCR was able to significantly reduce reconstruction errors, and subsequent functional activation errors, relative to SR at identical computational cost. Alternatively, at fixed reconstruction accuracy, computation time was greatly reduced. The improved performance was particularly obvious for L1-norm nonlinear reconstructions relative to L2-norm linear reconstructions and robust to different regularization strength, undersampling rates, and off-resonance effects intensity. By examining activation maps, tPCR was also found to give similar improvements in real fMRI experiments.

CONCLUSION

The proposed proof-of-concept tPCR framework could improve (1) the reconstruction efficiency of iterative reconstruction, and (2) the reconstruction stability especially for nonlinear reconstructions. As a practical consideration, the improved reconstruction speed promotes the application of highly undersampled non-Cartesian fast fMRI.

Temperature-Sensitive Frozen-Tissue Imaging for Cryoablation Monitoring Using STIR-UTE MRI

PURPOSE

The aim of this study was to develop a method to delineate the lethally frozen-tissue region (temperature less than -40°C) arising from interventional cryoablation procedures using a short tau inversion-recovery ultrashort echo-time (STIR-UTE) magnetic resonance (MR) imaging sequence. This method could serve as an intraprocedural validation of the completion of tumor ablation, reducing the number of local recurrences after cryoablation procedures.

MATERIALS AND METHODS

The method relies on the short T1 and T2* relaxation times of frozen soft tissue. Pointwise Encoding Time with Radial Acquisition, a 3-dimensional UTE sequence with TE = 70 microseconds, was optimized with STIR to null tissues with a T1 of approximately 271 milliseconds, the threshold T1. Because the T1 relaxation time of frozen tissue in the temperature range of -40°C < temperature < -8°C is shorter than the threshold T1 at the 3-tesla magnetic field, tissues in this range should appear hyperintense. The sequence was evaluated in ex vivo frozen tissue, where image intensity and actual tissue temperatures, measured by thermocouples, were correlated. Thereafter, the sequence was evaluated clinically in 12 MR-guided prostate cancer cryoablations, where MR-compatible cryoprobes were used to destroy cancerous tissue and preserve surrounding normal tissue.

RESULTS

The ex vivo experiment using a bovine muscle demonstrated that STIR-UTE images showed regions approximately between -40°C and -8°C as hyperintense, with tissues at lower and higher temperatures appearing dark, making it possible to identify the region likely to be above the lethal temperature inside the frozen tissue. In the clinical cases, the STIR-UTE images showed a dark volume centered on the cryoprobe shaft, Vinner, where the temperature is likely below -40°C, surrounded by a doughnut-shaped hyperintense volume, where the temperature is likely between -40°C and -8°C. The hyperintense region was itself surrounded by a dark volume, where the temperature is likely above -8°C, permitting calculation of Vouter. The STIR-UTE frozen-tissue volumes, Vinner and Vouter, appeared significantly smaller than signal voids on turbo spin echo images (P < 1.0 × 10), which are currently used to quantify the frozen-tissue volume ("the iceball"). The ratios of the Vinner and Vouter volumes to the iceball were 0.92 ± 0.08 and 0.29 ± 0.07, respectively. In a single postablation follow-up case, a strong correlation was seen between Vinner and the necrotic volume.

CONCLUSIONS

Short tau inversion-recovery ultrashort echo-time MR imaging successfully delineated the area approximately between -40°C and -8°C isotherms in the frozen tissue, demonstrating its potential to monitor the lethal ablation volume during MR-guided cryoablation.

A Network-Driven Prior Induced Bregman Model for Parallel MR Imaging

Compressed sensing based parallel imaging (CS-PI) has attracted great attention in fast magnetic resonance imaging (MRI) community. In particular, Bregman iterative model has shown encouraging performance in solving this problem. However, its regularization term still has large room for improvement. In this work, we propose a network-driven prior induced Bregman model, dubbed as Breg-EDAEP, for CS-PI task. In the present model, the implicit property among different channel MR images is preliminarily explored by the network to obtain more structure details in iterative reconstruction procedure. Experiments on various acceleration factors and sampling patterns have shown that the proposed method outperforms the state-of-the-art algorithms. Breg-EDAEP possesses strong capability to restore image details and preserves well structure information.

Deep-Learning Methods for Parallel Magnetic Resonance Imaging Reconstruction: A Survey of the Current Approaches, Trends, and Issues

Following the success of deep learning in a wide range of applications, neural network-based machine learning techniques have received interest as a means of accelerating magnetic resonance imaging (MRI). A number of ideas inspired by deep learning techniques from computer vision and image processing have been successfully applied to non-linear image reconstruction in the spirit of compressed sensing for both low dose computed tomography and accelerated MRI. The additional integration of multi-coil information to recover missing k-space lines in the MRI reconstruction process, is still studied less frequently, even though it is the de-facto standard for currently used accelerated MR acquisitions. This manuscript provides an overview of the recent machine learning approaches that have been proposed specifically for improving parallel imaging. A general background introduction to parallel MRI is given that is structured around the classical view of image space and k-space based methods. Both linear and non-linear methods are covered, followed by a discussion of recent efforts to further improve parallel imaging using machine learning, and specifically using artificial neural networks. Image-domain based techniques that introduce improved regularizers are covered as well as k-space based methods, where the focus is on better interpolation strategies using neural networks. Issues and open problems are discussed as well as recent efforts for producing open datasets and benchmarks for the community.

Overdiscrete echo-planar spectroscopic imaging with correlated higher-order phase correction

PURPOSE

To introduce a robust methodology for fast ¹ H MRSI of the brain at 3T with improved SNR and reduced phase-related artifacts.

METHOD

An accelerated acquisition scheme using echo-planar spectroscopic imaging (EPSI) was combined with the overdiscrete reconstruction framework. This approach enables the interleaved acquisition of a water reference scan at each phase encoding step, maximizing its correlation with the water-suppressed measurement. Moreover, a generalized high-order phase correction was incorporated into the reconstruction pipeline. The spatial-temporal phase correction term was estimated from the reference scan and interpolated to high resolution using a polynomial basis. The method was implemented at 3T and validated with phantom and in vivo experiments.

RESULTS

The methodology showed the elimination of spectral artifacts generated by phase disturbances and achieved mean SNR gains in vivo of 3.18 and 1.19 compared to standard reconstructions with corrections performed at nominal and high resolution, respectively. EPSI scans with interleaved water acquisition showed to be robust to system instabilities and potentially to patient motion. Moreover, phase distortions were effectively corrected in a single step, avoiding additional reference measurements and post-processing steps.

CONCLUSION

The overdiscrete reconstruction framework with high-order phase correction allowed to effectively correct for distortions, related to B₀ inhomogeneities, B₀ drift, eddy currents, and system vibrations. Furthermore, the presented reconstruction method, combined with EPSI acquisitions, demonstrated improved measurement stability, substantial SNR enhancement, better spectral linewidth, and effective artifact removal.

Comparison of different methods for the estimation of aortic pulse wave velocity from 4D flow cardiovascular magnetic resonance

BACKGROUND

Arterial pulse wave velocity (PWV) is associated with increased mortality in aging and disease. Several studies have shown the accuracy of applanation tonometry carotid-femoral PWV (Cf-PWV) and the relevance of evaluating central aorta stiffness using 2D cardiovascular magnetic resonance (CMR) to estimate PWV, and aortic distensibility-derived PWV through the theoretical Bramwell-Hill model (BH-PWV). Our aim was to compare various methods of aortic PWV (aoPWV) estimation from 4D flow CMR, in terms of associations with age, Cf-PWV, BH-PWV and left ventricular (LV) mass-to-volume ratio while evaluating inter-observer reproducibility and robustness to temporal resolution.

METHODS

We studied 47 healthy subjects (49.5 ± 18 years) who underwent Cf-PWV and CMR including aortic 4D flow CMR as well as 2D cine SSFP for BH-PWV and LV mass-to-volume ratio estimation. The aorta was semi-automatically segmented from 4D flow data, and mean velocity waveforms were estimated in 25 planes perpendicular to the aortic centerline. 4D flow CMR aoPWV was calculated: using velocity curves at two locations, namely ascending aorta (AAo) and distal descending aorta (DAo) aorta (S1, 2D-like strategy), or using all velocity curves along the entire aortic centreline (3D-like strategies) with iterative transit time (TT) estimates (S2) or a plane fitting of velocity curves systolic upslope (S3). For S1 and S2, TT was calculated using three approaches: cross-correlation (TTc), wavelets (TTw) and Fourier transforms (TTf). Intra-class correlation coefficients (ICC) and Bland-Altman biases (BA) were used to evaluate inter-observer reproducibility and effect of lower temporal resolution.

RESULTS

4D flow CMR aoPWV estimates were significantly (p < 0.05) correlated to the CMR-independent Cf-PWV, BH-PWV, age and LV mass-to-volume ratio, with the strongest correlations for the 3D-like strategy using wavelets TT (S2-TTw) (R = 0.62, 0.65, 0.77 and 0.52, respectively, all p < 0.001). S2-TTw was also highly reproducible (ICC = 0.99, BA = 0.09 m/s) and robust to lower temporal resolution (ICC = 0.97, BA = 0.15 m/s).

CONCLUSIONS

Reproducible 4D flow CMR aoPWV estimates can be obtained using full 3D aortic coverage. Such 4D flow CMR stiffness measures were significantly associated with Cf-PWV, BH-PWV, age and LV mass-to-volume ratio, with a slight superiority of the 3D strategy using wavelets transit time (S2-TTw).

A quantitative comparison between a navigated Cartesian and a self-navigated radial protocol from clinical studies for free-breathing 3D whole-heart bSSFP coronary MRA

PURPOSE

Navigator-gated 3D bSSFP whole-heart coronary MRA has been evaluated in several large studies including a multi-center trial. Patient studies have also been performed with more recent self-navigated techniques. In this study, these two approaches are compared side-by-side using a Cartesian navigator-gated and corrected (CNG) and a 3D radial self-navigated (RSN) protocol from published patient studies.

METHODS

Sixteen healthy subjects were examined with both sequences on a 1.5T scanner. Assessment of the visibility of coronary ostia and quantitative comparisons of acquisition times, blood pool homogeneity, and visible length and sharpness of the right coronary artery (RCA) and the combined left main (LM)+left anterior descending (LAD) coronary arteries were performed. Paired sample t-tests with P < .05 considered statistically significant were used for all comparisons.

RESULTS

The acquisition time was 5:40 ± 0:28 min (mean ± SD) for RSN, being significantly shorter than the 16:59 ± 5:05 min of CNG (P < .001). RSN images showed higher blood pool homogeneity (P < .001). All coronary ostia were visible with both techniques. CNG provided significantly higher vessel sharpness in the RCA (CNG: 50.0 ± 8.6%, RSN: 34.2 ± 6.9%, P < .001) and the LM+LAD (CNG: 48.7 ± 6.7%, RSN: 32.3 ± 7.1%, P < .001). The visible vessel length was significantly longer in the LM+LAD using CNG (CNG: 9.8 ± 2.7 cm, RSN: 8.5 ± 2.6 cm, P < .05) but not in the RCA (CNG: 9.7 ± 2.3 cm, RSN: 9.3 ± 2.9 cm, P = .29).

CONCLUSION

CNG provided superior vessel sharpness and might hence be the better option for examining coronary lumina. However, its blood pool inhomogeneity and prolonged and unpredictable acquisition times compared to RSN may make clinical adoption more challenging.

Improved Regularized Reconstruction for Simultaneous Multi-Slice Cardiac MRI T 1 Mapping

Myocardial T ₁ mapping is a quantitative MRI technique that has found great clinical utility in the detection of various heart disease. These acquisitions typically require three breath-holds, leading to long scan durations and patient discomfort. Simultaneous multi-slice (SMS) imaging has been shown to reduce the scan time of myocardial T ₁ mapping to a single breath-hold without sacrificing coverage, albeit at reduced precision. In this work, we propose a new reconstruction strategy for SMS imaging that combines the advantages of two different k-space interpolation strategies, while allowing for regularization, in order to improve the precision of accelerated mycordial T ₁ mapping.

Perturbed spiral real-time phase-contrast MR with compressive sensing reconstruction for assessment of flow in children

PURPOSE

we implemented a golden-angle spiral phase contrast sequence. A commonly used uniform density spiral and a new 'perturbed' spiral that produces more incoherent aliases were assessed. The aim was to ascertain whether greater incoherence enabled more accurate Compressive Sensing reconstruction and superior measurement of flow and velocity.

METHODS

A range of 'perturbed' spiral trajectories based on a uniform spiral trajectory were formulated. The trajectory that produced the most noise-like aliases was selected for further testing. For in-silico and in-vivo experiments, data was reconstructed using total Variation L1 regularisation in the spatial and temporal domains. In-silico, the reconstruction accuracy of the 'perturbed' golden spiral was compared to uniform density golden-angle spiral. For the in-vivo experiment, stroke volume and peak mean velocity were measured in 20 children using 'perturbed' and uniform density golden-angle spiral sequences. These were compared to a reference standard gated Cartesian sequence.

RESULTS

In-silico, the perturbed spiral acquisition produced more accurate reconstructions with less temporal blurring (NRMSE ranging from 0.03 to 0.05) than the uniform density acquisition (NRMSE ranging from 0.06 to 0.12). This translated in more accurate results in-vivo with no significant bias in the peak mean velocity (bias: -0.1, limits: -4.4 to 4.1 cm/s; P = 0.98) or stroke volume (bias: -1.8, limits: -9.4 to 5.8 ml, P = 0.19).

CONCLUSION

We showed that a 'perturbed' golden-angle spiral approach is better suited to Compressive Sensing reconstruction due to more incoherent aliases. This enabled accurate real-time measurement of flow and peak velocity in children.

Adenosine stress CMR perfusion imaging of the temporal evolution of perfusion defects in a porcine model of progressive obstructive coronary artery occlusion

Adenosine stress CMR perfusion imaging can quantify absolute perfusion and myocardial perfusion reserve (MPR) in coronary artery disease (CAD) with higher spatial resolution than positron emission tomography, the only clinically available technique for quantitative myocardial perfusion imaging. While porcine models of CAD are excellent for studying perfusion abnormalities in chronic CAD, to date there are a limited number of studies that use quantitative perfusion for evaluation. Therefore, we developed an adenosine stress CMR protocol to evaluate the temporal evolution of perfusion defects in a porcine model of progressive obstructive CAD. 10 Yucatan minipigs underwent placement of an ameroid occluder around the left circumflex artery (LCX) to induce a progressive chronic coronary obstruction. Four animals underwent a hemodynamic dose range experiment to determine the adenosine dose inducing maximal hyperemia. Each animal had a CMR examination, including stress/rest spiral quantitative perfusion imaging at baseline and 1, 3, and 6 weeks. Late gadolinium enhancement images determined the presence of myocardial infarction, if any existed. Pixelwise quantitative perfusion maps were generated using Fermi deconvolution. The results were statistically analyzed with a repeated mixed measures model to block for physiological variation between the animals. Five animals developed myocardial infarction by 3 weeks, while three developed ischemia without an infarction. The perfusion defects were located in the inferolateral myocardium in the perfusion territory of the LCX. Stress perfusion values were higher in remote segments than both the infarcted and ischemic segments (p < 0.01). MPR values were significantly greater in the remote segments than infarcted and ischemic segments (p < 0.01). While the MPR decreased in all segments, the MPR recovered by the sixth week in the remote regions. We developed a model of progressive CAD and evaluated the temporal evolution of the development of quantitative perfusion defects. This model will serve as a platform for understanding the development of perfusion abnormalities in chronic occlusive CAD.

SANTIS: Sampling-Augmented Neural neTwork with Incoherent Structure for MR image reconstruction

PURPOSE

To develop and evaluate a novel deep learning-based reconstruction framework called SANTIS (Sampling-Augmented Neural neTwork with Incoherent Structure) for efficient MR image reconstruction with improved robustness against sampling pattern discrepancy.

METHODS

With a combination of data cycle-consistent adversarial network, end-to-end convolutional neural network mapping, and data fidelity enforcement for reconstructing undersampled MR data, SANTIS additionally utilizes a sampling-augmented training strategy by extensively varying undersampling patterns during training, so that the network is capable of learning various aliasing structures and thereby removing undersampling artifacts more effectively and robustly. The performance of SANTIS was demonstrated for accelerated knee imaging and liver imaging using a Cartesian trajectory and a golden-angle radial trajectory, respectively. Quantitative metrics were used to assess its performance against different references. The feasibility of SANTIS in reconstructing dynamic contrast-enhanced images was also demonstrated using transfer learning.

RESULTS

Compared to conventional reconstruction that exploits image sparsity, SANTIS achieved consistently improved reconstruction performance (lower errors and greater image sharpness). Compared to standard learning-based methods without sampling augmentation (e.g., training with a fixed undersampling pattern), SANTIS provides comparable reconstruction performance, but significantly improved robustness, against sampling pattern discrepancy. SANTIS also achieved encouraging results for reconstructing liver images acquired at different contrast phases.

CONCLUSION

By extensively varying undersampling patterns, the sampling-augmented training strategy in SANTIS can remove undersampling artifacts more robustly. The novel concept behind SANTIS can particularly be useful for improving the robustness of deep learning-based image reconstruction against discrepancy between training and inference, an important, but currently less explored, topic.

Applications of magnetic resonance imaging in chemical engineering

While there are many techniques to study phenomena that occur in chemical engineering applications, magnetic resonance imaging (MRI) receives increasing scientific interest. Its non-invasive nature and wealth of parameters with the ability to generate functional images and contrast favors the use of MRI for many purposes, in particular investigations of dynamic phenomena, since it is very sensitive to motion. Recent progress in flow-MRI has led to shorter acquisition times and enabled studies of transient phenomena. Reactive systems can easily be imaged if NMR parameters such as relaxation change along the reaction coordinate. Moreover, materials and devices can be examined, such as batteries by mapping the magnetic field around them.

Self-calibrating wave-encoded 3D turbo spin echo imaging using subspace model based autofocusing

PURPOSE

To develop a self-calibrating approach for the estimation of wave point spread function (PSF) and coil sensitivities from the subsampled wave-encoded k-space, and evaluate its performance for wave-encoded 3D turbo spin echo (TSE) imaging.

METHODS

A low rank subspace parametric model was demonstrated in simulation to improve the representation for practical wave encoding k-space trajectories with aperiodicity, and an autofocus metric for the entire imaging volume was used to calibrate the wave PSF in a 2-stage manner from coarse to refined estimation. The coil sensitivities can be extracted from the shifted central region of wave PSF corrected subsampled k-space, and further used with wave PSF for wave-encoded parallel imaging (PI) reconstruction. The wave encoding gradients were integrated into the 3D TSE sequence considering eddy current reduction aspects and maintaining of the Carr-Purcell-Meiboom-Gill condition. Phantom and in vivo brain experiments were performed to evaluate the accuracy of wave PSF self-calibration and to compare the PI reconstruction performance between wave and Cartesian encoding scheme.

RESULTS

The self-calibrated wave PSF, estimated from different k-space undersampling patterns can robustly correct the wave encoding induced image artifacts given sufficient central autocalibration data. The self-calibrating wave-encoded PI reconstruction has demonstrated its improved performance in reduced aliasing artifacts and noise amplification in comparison to the Cartesian-encoded PI reconstruction results for 3D TSE imaging.

CONCLUSION

The proposed self-calibrating wave-encoded method allows robust calibration of wave PSF and coil sensitivities from the subsampled k-space, and improves the overall image quality for accelerated 3D TSE imaging.