Assessment of deep learning segmentation for real-time free-breathing cardiac magnetic resonance imaging at rest and under exercise stress

In recent years, a variety of deep learning networks for cardiac MRI (CMR) segmentation have been developed and analyzed. However, nearly all of them are focused on cine CMR under breathold. In this work, accuracy of deep learning methods is assessed for volumetric analysis (via segmentation) of the left ventricle in real-time free-breathing CMR at rest and under exercise stress. Data from healthy volunteers (n = 15) for cine and real-time free-breathing CMR at rest and under exercise stress were analyzed retrospectively. Exercise stress was performed using an ergometer in the supine position. Segmentations of two deep learning methods, a commercially available technique (comDL) and an openly available network (nnU-Net), were compared to a reference model created via the manual correction of segmentations obtained with comDL. Segmentations of left ventricular endocardium (LV), left ventricular myocardium (MYO), and right ventricle (RV) are compared for both end-systolic and end-diastolic phases and analyzed with Dice's coefficient. The volumetric analysis includes the cardiac function parameters LV end-diastolic volume (EDV), LV end-systolic volume (ESV), and LV ejection fraction (EF), evaluated with respect to both absolute and relative differences. For cine CMR, nnU-Net and comDL achieve Dice's coefficients above 0.95 for LV and 0.9 for MYO, and RV. For real-time CMR, the accuracy of nnU-Net exceeds that of comDL overall. For real-time CMR at rest, nnU-Net achieves Dice's coefficients of 0.94 for LV, 0.89 for MYO, and 0.90 for RV and the mean absolute differences between nnU-Net and the reference are 2.9 mL for EDV, 3.5 mL for ESV, and 2.6% for EF. For real-time CMR under exercise stress, nnU-Net achieves Dice's coefficients of 0.92 for LV, 0.85 for MYO, and 0.83 for RV and the mean absolute differences between nnU-Net and reference are 11.4 mL for EDV, 2.9 mL for ESV, and 3.6% for EF. Deep learning methods designed or trained for cine CMR segmentation can perform well on real-time CMR. For real-time free-breathing CMR at rest, the performance of deep learning methods is comparable to inter-observer variability in cine CMR and is usable for fully automatic segmentation. For real-time CMR under exercise stress, the performance of nnU-Net could promise a higher degree of automation in the future.

Accelerated 2D Cartesian MRI with an 8-channel local B0 coil array combined with parallel imaging

PURPOSE

In MRI, the magnetization of nuclear spins is spatially encoded with linear gradients and radiofrequency receivers sensitivity profiles to produce images, which inherently leads to a long scan time. Cartesian MRI, as widely adopted for clinical scans, can be accelerated with parallel imaging and rapid magnetic field modulation during signal readout. Here, by using an 8-channel localB 0 $$ {\mathrm{B}}_0 $$ coil array, the modulation scheme optimized for sampling efficiency is investigated to speed up 2D Cartesian scans.

THEORY AND METHODS

An 8-channel localB 0 $$ {\mathrm{B}}_0 $$ coil array is made to carry sinusoidal currents during signal readout to accelerate 2D Cartesian scans. An MRI sampling theory based on reproducing kernel Hilbert space is exploited to visualize the efficiency of nonlinear encoding in arbitrary sampling duration. A field calibration method using current monitors for localB 0 $$ {\mathrm{B}}_0 $$ coils and the ESPIRiT algorithm is proposed to facilitate image reconstruction. Image acceleration with various modulation field shapes, aliasing control, and distinct modulation frequencies are scrutinized to find an optimized modulation scheme. A safety evaluation is conducted. In vivo 2D Cartesian scans are accelerated by the localB 0 $$ {\mathrm{B}}_0 $$ coils.

RESULTS

For 2D Cartesian MRI, the optimal modulation field by this localB 0 $$ {\mathrm{B}}_0 $$ array converges to a nearly linear gradient field. With the field calibration technique, it accelerates the in vivo scans (i.e., proved safe) by threefold and eightfold free of visible artifacts, without and with SENSE, respectively.

CONCLUSION

The nonlinear encoding analysis tool, the field calibration method, the safety evaluation procedures, and the in vivo reconstructed scans make significant steps to push MRI speed further with the localB 0 $$ {\mathrm{B}}_0 $$ coil array.

Rational Approximation of Golden Angles for Simple and Reproducible Radial Sampling

PURPOSE

To develop a generic radial sampling scheme that combines the advantages of golden ratio sampling with simplicity of equidistant angular patterns. The irrational angle between consecutive spokes in golden ratio based sampling schemes enables a flexible retrospective choice of temporal resolution, while preserving good coverage of k-space for each individual bin. Nevertheless, irrational increments prohibit precomputation of the point-spread function (PSF), can lead to numerical problems, and require more complex processing steps. To avoid these problems, a new sampling scheme based on a rational approximation of golden angles (RAGA) is developed.

METHODS

The theoretical properties of RAGA sampling are mathematically derived. Sidelobe-to-peak ratios (SPR) are numerically computed and compared to the corresponding golden ratio sampling schemes. The sampling scheme is implemented in the BART toolbox and in a radial gradient-echo sequence. Feasibility is shown for quantitative imaging in a phantom and a cardiac scan of a healthy volunteer.

RESULTS

RAGA sampling can accurately approximate golden ratio sampling and has almost identical PSF and SPR. In contrast to golden ratio sampling, each frame can be reconstructed with the same equidistant trajectory using different sampling masks, and the angle of each acquired spoke can be encoded as a small index, which simplifies processing of the acquired data.

CONCLUSION

RAGA sampling provides the advantages of golden ratio sampling while simplifying data processing, rendering it a valuable tool for dynamic and quantitative MRI.

Visualization of deglutition and gastroesophageal reflux using real-time MRI: a standardized approach to image acquisition and assessment

This study aims to develop a standardized algorithm for gastroesophageal image acquisition and diagnostic assessment using real-time MRI. Patients with GERD symptoms undergoing real-time MRI of the esophagus and esophagogastric junction between 2015 and 2018 were included. A 10 ml bolus of pineapple juice served as an oral contrast agent. Patients performed Valsalva maneuver to provoke reflux and hiatal hernia. Systematic MRI assessment included visual presence of achalasia, fundoplication failure in patients with previous surgical fundoplication, gastroesophageal reflux, and hiatal hernia. A total of 184 patients (n = 92 female [50%], mean age 52.7 ± 15.8 years) completed MRI studies without adverse events at a mean examination time of 15 min. Gastroesophageal reflux was evident in n = 117 (63.6%), hiatal hernia in n = 95 (52.5%), and achalasia in 4 patients (2.2%). Hiatal hernia was observed more frequently in patients with reflux at rest (n = 67 vs. n = 6, p < 0.01) and during Valsalva maneuver (n = 87 vs. n = 8, p < 0.01). Real-time MRI visualized a morphologic correlate for recurring GERD symptoms in 20/22 patients (90%) after fundoplication procedure. In a large-scale single-center cohort of patients with GERD symptoms undergoing real-time MRI, visual correlates for clinical symptoms were evident in most cases. The proposed assessment algorithm could aid in wider-spread utilization of real-time MRI and provides a comprehensive approach to this novel imaging modality.

Evaluation of MR-safe bioptomes for MR-guided endomyocardial biopsy in minipigs: a potential radiation-free clinical approach

BACKGROUND

Diagnostic accuracy of endomyocardial biopsy could improve if clinically safe magnetic resonance (MR)-compatible bioptomes were available. We explored two novel MR-compatible cardiac bioptomes for performance, safety, and clinical viability, employing in vivo minipig trials and phase-contrast synchrotron radiation computed microtomography (SRµCT).

METHODS

Analysis of ex vivo obtained pig endomyocardial biopsies was performed using phase-contrast SRµCT and conventional two-dimensional histology. The technical performance was evaluated by measuring volume, inner and outer integrities, compression, and histological diagnostic value in 3 sets (6 per set) of biopsies for each experimental bioptome. The bioptomes were tested in vivo in 3 healthy minipigs per bioptome. The clinical feasibility was evaluated by procedural and cutting success as well as histological diagnostic value.

RESULTS

The bioptome with the 'grind-grind' design achieved similar values to control in compression (p = 0.822), inner (p = 0.628), and outer (p = 0.507), integrities ex vivo. It showed a better performance in the in vivo real-time MRI setting demonstrating a higher cutting success (91.7%) than the 'grind-anvil' (86.2%) design. In both ex vivo and in vivo evaluations, the 'grind-grind' design displayed sufficient diagnostic value (83% and 95%). The 'grind-anvil' design showed adequate diagnostic value both ex vivo and in vivo (78% and 87.5%) but was not comparable to control according to the three-dimensional (3D) analysis.

CONCLUSION

A novel MR-compatible bioptome was identified as plausible in a clinical setting. Additionally, SRµCT and subsequent 3D structural analysis could be valuable in the label-free investigation of myocardial tissue at a micrometer level.

RELEVANCE STATEMENT

Implementation of MR-guided biopsy can improve animal studies on structural myocardial changes at any point in an experimental setup. With further improvements in guiding catheters, MR-guided biopsy, using the new bioptome, has a potential to increase quality and diagnostic accuracy in patients both with structural and inflammatory cardiomyopathies.

KEY POINTS

• Novel MR-compatible bioptomes show promise for a clinical application. • SRµCT enabled detailed analysis of endomyocardial biopsies. • The bioptomes showed adequate in vivo performance without major complications.

Free-Breathing Liver Fat, R₂* and B₀ Field Mapping Using Multi-Echo Radial FLASH and Regularized Model-Based Reconstruction

This work introduced a stack-of-radial multi-echo asymmetric-echo MRI sequence for free-breathing liver volumetric acquisition. Regularized model-based reconstruction was implemented in Berkeley Advanced Reconstruction Toolbox (BART) to jointly estimate all physical parameter maps (water, fat, R2∗ , and B0 field inhomogeneity maps) and coil sensitivity maps from self-gated k -space data. Specifically, locally low rank and temporal total variation regularization were employed directly on physical parameter maps. The proposed free-breathing radial technique was tested on a water/fat & iron phantom, a young volunteer, and obesity/diabetes/hepatic steatosis patients. Quantitative fat fraction and R2∗ accuracy were confirmed by comparing our technique with the reference breath-hold Cartesian scan. The multi-echo radial sampling sequence achieves fast k -space coverage and is robust to motion. Moreover, the proposed motion-resolved model-based reconstruction allows for free-breathing liver fat and R2∗ quantification in multiple motion states. Overall, our proposed technique offers a convenient tool for non-invasive liver assessment with no breath holding requirement.

Quantitative MRI by nonlinear inversion of the Bloch equations

PURPOSE

Development of a generic model-based reconstruction framework for multiparametric quantitative MRI that can be used with data from different pulse sequences.

METHODS

Generic nonlinear model-based reconstruction for quantitative MRI estimates parametric maps directly from the acquired k-space by numerical optimization. This requires numerically accurate and efficient methods to solve the Bloch equations and their partial derivatives. In this work, we combine direct sensitivity analysis and pre-computed state-transition matrices into a generic framework for calibrationless model-based reconstruction that can be applied to different pulse sequences. As a proof-of-concept, the method is implemented and validated for quantitative T 1 $$ {\mathrm{T}}_1 $$ and T 2 $$ {\mathrm{T}}_2 $$ mapping with single-shot inversion-recovery (IR) FLASH and IR bSSFP sequences in simulations, phantoms, and the human brain.

RESULTS

The direct sensitivity analysis enables a highly accurate and numerically stable calculation of the derivatives. The state-transition matrices efficiently exploit repeating patterns in pulse sequences, speeding up the calculation by a factor of 10 for the examples considered in this work, while preserving the accuracy of native ordinary differential equations solvers. The generic model-based method reproduces quantitative results of previous model-based reconstructions based on the known analytical solutions for radial IR FLASH. For IR bSFFP it produces accurate T 1 $$ {\mathrm{T}}_1 $$ and T 2 $$ {\mathrm{T}}_2 $$ maps for the National Insitute of Standards and Technology (NIST) phantom in numerical simulations and experiments. Feasibility is also shown for human brain, although results are affected by magnetization transfer effects.

CONCLUSION

By developing efficient tools for numerical optimizations using the Bloch equations as forward model, this work enables generic model-based reconstruction for quantitative MRI.

Free-breathing myocardial T1 mapping using inversion-recovery radial FLASH and motion-resolved model-based reconstruction

PURPOSE

To develop a free-breathing myocardialT 1 $$ {\mathrm{T}}_1 $$ mapping technique using inversion-recovery (IR) radial fast low-angle shot (FLASH) and calibrationless motion-resolved model-based reconstruction.

METHODS

Free-running (free-breathing, retrospective cardiac gating) IR radial FLASH is used for data acquisition at 3T. First, to reduce the waiting time between inversions, an analytical formula is derived that takes the incompleteT 1 $$ {\mathrm{T}}_1 $$ recovery into account for an accurateT 1 $$ {\mathrm{T}}_1 $$ calculation. Second, the respiratory motion signal is estimated from the k-space center of the contrast varying acquisition using an adapted singular spectrum analysis (SSA-FARY) technique. Third, a motion-resolved model-based reconstruction is used to estimate both parameter and coil sensitivity maps directly from the sorted k-space data. Thus, spatiotemporal total variation, in addition to the spatial sparsity constraints, can be directly applied to the parameter maps. Validations are performed on an experimental phantom, 11 human subjects, and a young landrace pig with myocardial infarction.

RESULTS

In comparison to an IR spin-echo reference, phantom results confirm goodT 1 $$ {\mathrm{T}}_1 $$ accuracy, when reducing the waiting time from 5 s to 1 s using the new correction. The motion-resolved model-based reconstruction further improvesT 1 $$ {\mathrm{T}}_1 $$ precision compared to the spatial regularization-only reconstruction. Aside from showing that a reliable respiratory motion signal can be estimated using modified SSA-FARY, in vivo studies demonstrate that dynamic myocardialT 1 $$ {\mathrm{T}}_1 $$ maps can be obtained within 2 min with good precision and repeatability.

CONCLUSION

Motion-resolved myocardialT 1 $$ {\mathrm{T}}_1 $$ mapping during free-breathing with good accuracy, precision and repeatability can be achieved by combining inversion-recovery radial FLASH, self-gating and a calibrationless motion-resolved model-based reconstruction.

Bayesian MRI reconstruction with joint uncertainty estimation using diffusion models

PURPOSE

We introduce a framework that enables efficient sampling from learned probability distributions for MRI reconstruction.

METHOD

Samples are drawn from the posterior distribution given the measured k-space using the Markov chain Monte Carlo (MCMC) method, different from conventional deep learning-based MRI reconstruction techniques. In addition to the maximum a posteriori estimate for the image, which can be obtained by maximizing the log-likelihood indirectly or directly, the minimum mean square error estimate and uncertainty maps can also be computed from those drawn samples. The data-driven Markov chains are constructed with the score-based generative model learned from a given image database and are independent of the forward operator that is used to model the k-space measurement.

RESULTS

We numerically investigate the framework from these perspectives: (1) the interpretation of the uncertainty of the image reconstructed from undersampled k-space; (2) the effect of the number of noise scales used to train the generative models; (3) using a burn-in phase in MCMC sampling to reduce computation; (4) the comparison to conventional ℓ 1 $$ {\ell}_1 $$ -wavelet regularized reconstruction; (5) the transferability of learned information; and (6) the comparison to fastMRI challenge.

CONCLUSION

A framework is described that connects the diffusion process and advanced generative models with Markov chains. We demonstrate its flexibility in terms of contrasts and sampling patterns using advanced generative priors and the benefits of also quantifying the uncertainty for every pixel.

Deep, deep learning with BART

PURPOSE

To develop a deep-learning-based image reconstruction framework for reproducible research in MRI.

METHODS

The BART toolbox offers a rich set of implementations of calibration and reconstruction algorithms for parallel imaging and compressed sensing. In this work, BART was extended by a nonlinear operator framework that provides automatic differentiation to allow computation of gradients. Existing MRI-specific operators of BART, such as the nonuniform fast Fourier transform, are directly integrated into this framework and are complemented by common building blocks used in neural networks. To evaluate the use of the framework for advanced deep-learning-based reconstruction, two state-of-the-art unrolled reconstruction networks, namely the Variational Network and MoDL, were implemented.

RESULTS

State-of-the-art deep image-reconstruction networks can be constructed and trained using BART's gradient-based optimization algorithms. The BART implementation achieves a similar performance in terms of training time and reconstruction quality compared to the original implementations based on TensorFlow.

CONCLUSION

By integrating nonlinear operators and neural networks into BART, we provide a general framework for deep-learning-based reconstruction in MRI.

Concomitant latent pulmonary vascular disease leads to impaired global cardiac performance in heart failure with preserved ejection fraction

AIMS

The REDUCE-LAP II trial demonstrated adverse outcomes after interatrial shunt device (IASD) placement in heart failure with preserved ejection fraction (HFpEF) attributed to latent pulmonary vascular disease (PVD). We hypothesized that exercise stress cardiovascular magnetic resonance (CMR) imaging could provide non-invasive characterization of cardiac and pulmonary physiology for improved patient selection.

METHODS AND RESULTS

The HFpEF-Stress trial prospectively enrolled 75 patients with exertional dyspnoea and diastolic dysfunction. Patients underwent rest and exercise stress right heart catheterization, echocardiography and CMR imaging. Pulmonary artery and capillary wedge pressures, cardiac index (CI) and pulmonary vascular resistance (PVR) were calculated. Latent PVD was defined as increased PVR ≥ 1.74 Wood units during exercise stress. CMR assessed long-axis strains (LAS) and filling volumes of all cardiac chambers. Right ventricular (RV) function was further quantified by stroke and peak flow volumes. Patients with latent PVD (n = 24) showed lower RV function (rest tricuspid annular plane systolic excursion, p = 0.010; stress RV LAS, p < 0.001) compared to patients without (n = 43). During exercise stress, RV stroke and peak flow volumes (p < 0.001) were reduced and led to impaired left atrial filling (p = 0.040) with a strong statistical trend to impaired ventricular (LV) filling (p = 0.098). This subsequently resulted in reduced LV-CI (p < 0.001) despite preserved LV systolic function (LV LAS p ≥ 0.255). The degree of RV dysfunction during exercise stress best predicted latent PVD (RV peak flow, area under the curve at rest 0.73 vs. stress 0.89, p = 0.004).

CONCLUSIONS

Latent PVD is a feature of HFpEF and is associated with impaired RV functional reserve, global diastolic filling and LV-CI. This can be quantified by CMR and used to identify patients likely to benefit from IASD implantation.

Real-time radial tagging for quantification of left ventricular torsion

PURPOSE

To develop a real-time radial tagging MRI for accurate measurement of rotational motion and twist of the left ventricle (LV).

METHODS

A FLASH-based radial tagging sequence with an undersampled radial reading scheme was developed for both single and double-slice imaging in real-time. The Polar Fourier Transform was used for reconstruction to push the undersampling artifacts out of a reduced FOV. The developed technique was used to image five normal subjects during rest, plus one during both exercise and rest conditions. LV rotational motions were estimated for five consecutive cardiac cycles in all cases. The process was validated using a numerical phantom. The real-time measurement of global rotational motion was compared with those measured from a non-real-time exam using linear regression analysis and the Bland-Altman plot.

RESULTS

The real-time acquisition was performed successfully with a temporal resolution of 46.2 ms. Image quality was sufficient for the reproducible calculation of rotation at rest and exercise. The feasibility of double-slice acquisition on human was further studied and a real-time twist of the left ventricle was demonstrated. The difference between LV global rotations from real-time and non-real-time approaches was 0.27 degrees. A significant reverse recoiling, induced by exercise, was reproducibly measured by the technique.

CONCLUSION

A real-time radial tagging MRI technique was developed based on the undersampled radial acquisition and Polar Fourier Transform reconstruction, for accurate measuring of the heart rotational motion and twist. The technique was able to extract a meaningful change of diastolic recoiling under stress test conditions during physical activities (cycling).

Imaging of arrhythmia: Real-time cardiac magnetic resonance imaging in atrial fibrillation

Objectives

Quantitative evaluations of function, volume and mass are fundamental in the diagnostic workup of different cardiovascular diseases and can be exactly determined by CMRI in sinus rhythm. This does not hold true in arrhythmia as CMR is hampered by reconstruction artifacts caused by inconsistent data from multiple heartbeats. Real-time (RT) MRI at high temporal resolution might reduce these problems.

Methods

Consecutive patients with atrial fibrillation were prospectively included and underwent RT and conventional CINE CMR in randomized order. 29 patients were studied at 1.5 T and 30 patients at 3 T. At 3 T a group of 20 subjects in sinus rhythm served as controls. RT and CINE image quality was evaluated in different planes and for different wall sections using a Likert scale (from zero to four). Volumetric analysis was performed using two types of software and differences between RT and CINE CMR were evaluated.

Results

In patients with atrial fibrillation RT CMR short axis (SA) resulted in a significantly higher image quality compared to CINE imaging both at 1.5 T and 3 T (1.5 T: mid SA: 3.55 ± 0.5 RT vs 2.6 ± 0.9 CINE, p = 0.0001; 3 T: mid SA: 3.15 ± 0.9 RT vs 2.6 ±1.0 CINE, p = 0.03); This qualitative difference was more marked and significant for the long axis views (2CV and 4CV) at 1.5 T (1.5 T: 2CV: 3.2 ± 0.6 RT vs 2.65 ± 1.1 CINE; p = 0.011; 4CV: 2.9 ± 0.69 RT vs 2.4 ± 0.9 CINE; p = 0.0044). During sinus rhythm CINE images were superior concerning diagnostic quality (3 T mid SA: 3.35 ± 0.45 RT vs 3.8 ± 0.5 CINE, p = 0.008). Quantitative analysis was successful with both software packages and the results showed a good correlation (Pearson correlation between 0.679 and 0.921 for patients). RT CMR resulted in slightly lower functional volumes than CINE CMR (3 T: patients: EDVI 86 ± 29 ml/m² RT vs 93 29 ml/m²± 29 CINE, Pearson r = 0.902) but similar ejection fractions (3 T: patients: EF 47 ± 16% RT vs 45 ± 13% CINE, Pearson r = 0679; controls: EF 63 ± 6 RT vs 63 ± 3 CINE, Pearson r = 0.695).

Conclusion

RT CMR improves image quality in arrhythmic patients and renders studies more comfortable. Volumetric analysis is feasible with slightly lower values relative to CINE CMR, while ejection fractions are comparable.

•

Real time cardiac magnetic imaging is superior to conventional CINE in arrhythmias: concerning image quality. Volumetric and functional analysis of real time is comparable to CINE. Acquistion time is reduced in real time. Improvement of postprocessing software of real time imaging is mandatory.

Assessment of esophagogastric junction morphology by dynamic real-time MRI: comparison of imaging features to high-resolution manometry

Purpose

To assess the esophagogastric junction (EGJ) on real-time MRI and compare imaging parameters to EGJ morphology on high-resolution manometry (HRM).

Methods

A total of 105 of 117 eligible patients who underwent real-time MRI and high-resolution manometry for GERD-like symptoms between 2015 and 2018 at a single center were retrospectively evaluated (male n = 57; female n = 48; mean age 52.5 ± 15.4 years). Real-time MRI was performed at a median investigation time of 15 min (1 frame/40 ms). On HRM, EGJ morphology was assessed according to the Chicago classification of esophageal motility disorders. Real-time MRI was performed at 3 T using highly undersampled radial fast low-angle shot acquisitions with NLINV image reconstruction. A 10 mL pineapple juice bolus served as oral contrast agent at supine position. Real-time MRI films of the EGJ were acquired during swallowing events and during Valsalva maneuver. Anatomic and functional MRI parameters were compared to EGJ morphology on HRM.

Results

On HRM, n = 42 patients presented with EGJ type I (40.0%), n = 33 with EGJ type II (31.4%), and n = 30 with EGJ type III (28.6%). On real-time MRI, hiatal hernia was more common in patients with EGJ type III (66.7%) than in patients with EGJ type I (26.2%) and EGJ type II (30.3%; p < 0.001). Sliding hiatal hernia was more frequent in patients with EGJ type II (33.3%) than in patients with EGJ type III (16.7%) and EGJ type I (7.1%; p = 0.017). The mean esophagus–fundus angle of patients was 85 ± 31° at rest and increased to 101 ± 36° during Valsalva maneuver.

Conclusion

Real-time MRI is a non-invasive imaging method for assessment of the esophagogastric junction. Real-time MRI can visualize dynamic changes of the EGJ during swallowing events.

Supplementary Information

The online version contains supplementary material available at 10.1007/s11604-021-01210-9.

Impaired Exercise Tolerance in Repaired Tetralogy of Fallot Is Associated With Impaired Biventricular Contractile Reserve: An Exercise-Stress Real-Time Cardiovascular Magnetic Resonance Study

BACKGROUND

Correction of tetralogy of Fallot (cTOF) often results in pulmonary valve pathology and right ventricular (RV) dysfunction. Reduced exercise capacity in cTOF patients cannot be explained by these findings alone. We aimed to explore why cTOF patients exhibit impaired exercise capacity with the aid of a comprehensive cardiopulmonary exercise testing (CPET) and real-time cardiovascular magnetic resonance exercise testing (CMR-ET) protocol.

METHODS

Thirty three cTOF patients and 35 matched healthy controls underwent CPET and CMR-ET in a prospective case-control study. Real-time steady-state free precession cine and phase-contrast sequences were obtained during incremental supine in-scanner cycling at 50, 70, and 90 W. RV and left ventricle (LV) volumes and pulmonary blood flow (Qp) were calculated. Differences of CPET and CMR-ET between cTOF versus controls and correlations between CPET and CMR-ET parameters in cTOF were evaluated statistically for all CMR exercise levels using Mann-Whitney U and Spearman rank-order correlation tests.

RESULTS

CPET capacity was significantly lower in cTOF than in controls. cTOF patients exhibited not only significantly reduced Qp and RV function but also lower LV function on CMR-ET. Higher CPET values in cTOF correlated with higher Qp (Qp 90 W versus carbon dioxide ventilatory equivalent %: R=-0.519, P<0.05), higher LV-end-diastolic volume indexed to body surface area (LV-end-diastolic volume indexed to body surface area at 50 W versus oxygen uptake in % at maximum exercise on CPET R=0.452, P<0.05), and change in LV ejection fraction (EF; LV-EF at 90 W versus Watt %: r=-0.463, P<0.05). No correlation was found with regard to RV-EF. Significant RV-LV interaction was observed during CMR-ET (RV-EF versus LV-EF at 50 W and 70 W: r=0.66, P<0.02 and r=0.52, P<0.05, respectively).

CONCLUSIONS

Impaired exercise capacity in cTOF resulted from a reduction in not only RV, but also LV function. cTOF with good exercise capacity on CPET demonstrated higher LV reserve and pulmonary blood flow during incremental CMR-ET. Apart from RV parameters, CMR-ET-derived LV function could be a valuable tool to stratify cTOF patients for pulmonary valve replacement.

Physics-based reconstruction methods for magnetic resonance imaging

Conventional magnetic resonance imaging (MRI) is hampered by long scan times and only qualitative image contrasts that prohibit a direct comparison between different systems. To address these limitations, model-based reconstructions explicitly model the physical laws that govern the MRI signal generation. By formulating image reconstruction as an inverse problem, quantitative maps of the underlying physical parameters can then be extracted directly from efficiently acquired k-space signals without intermediate image reconstruction-addressing both shortcomings of conventional MRI at the same time. This review will discuss basic concepts of model-based reconstructions and report on our experience in developing several model-based methods over the last decade using selected examples that are provided complete with data and code. This article is part of the theme issue 'Synergistic tomographic image reconstruction: part 1'.

Model-based reconstruction for simultaneous multi-slice T1 mapping using single-shot inversion-recovery radial FLASH

PURPOSE

To develop a single-shot multi-slice T 1 mapping method by combing simultaneous multi-slice (SMS) excitations, single-shot inversion-recovery (IR) radial fast low-angle shot (FLASH), and a nonlinear model-based reconstruction method.

METHODS

SMS excitations are combined with a single-shot IR radial FLASH sequence for data acquisition. A previously developed single-slice calibrationless model-based reconstruction is extended to SMS, formulating the estimation of parameter maps and coil sensitivities from all slices as a single nonlinear inverse problem. Joint-sparsity constraints are further applied to the parameter maps to improve T 1 precision. Validations of the proposed method are performed for a phantom and for the human brain and liver in 6 healthy adult subjects.

RESULTS

Phantom results confirm good T 1 accuracy and precision of the simultaneously acquired multi-slice T 1 maps in comparison to single-slice references. In vivo human brain studies demonstrate the better performance of SMS acquisitions compared to the conventional spoke-interleaved multi-slice acquisition using model-based reconstruction. Aside from good accuracy and precision, the results of 6 healthy subjects in both brain and abdominal studies confirm good repeatability between scan and re-scans. The proposed method can simultaneously acquire T 1 maps for 5 slices of a human brain ( 0.75 × 0.75 × 5 mm 3 ) or 3 slices of the abdomen ( 1.25 × 1.25 × 6 mm 3 ) within 4 seconds.

CONCLUSIONS

The IR SMS radial FLASH acquisition together with a nonlinear model-based reconstruction enable rapid high-resolution multi-slice T 1 mapping with good accuracy, precision, and repeatability.

Exercise Stress Real-Time Cardiac Magnetic Resonance Imaging for Noninvasive Characterization of Heart Failure With Preserved Ejection Fraction: The HFpEF-Stress Trial

BACKGROUND

Right heart catheterization using exercise stress is the reference standard for the diagnosis of heart failure with preserved ejection fraction (HFpEF) but carries the risk of the invasive procedure. We hypothesized that real-time cardiac magnetic resonance (RT-CMR) exercise imaging with pathophysiologic data at excellent temporal and spatial resolution may represent a contemporary noninvasive alternative for diagnosing HFpEF.

METHODS

The HFpEF-Stress trial (CMR Exercise Stress Testing in HFpEF; URL: https://www.clinicaltrials.gov; Unique identifier: NCT03260621. URL: https://dzhk.de/; Unique identifier: DZHK-17) prospectively recruited 75 patients with echocardiographic signs of diastolic dysfunction and dyspnea on exertion (E/e'>8, New York Heart Association class ≥II) to undergo echocardiography, right heart catheterization, and RT-CMR at rest and during exercise stress. HFpEF was defined according to pulmonary capillary wedge pressure (≥15 mm Hg at rest or ≥25 mm Hg during exercise stress). RT-CMR functional assessments included time-volume curves for total and early (1/3) diastolic left ventricular filling, left atrial (LA) emptying, and left ventricular/LA long axis strain.

RESULTS

Patients with HFpEF (n=34; median pulmonary capillary wedge pressure at rest, 13 mm Hg; at stress, 27 mm Hg) had higher E/e' (12.5 versus 9.15), NT-proBNP (N-terminal pro-B-type natriuretic peptide; 255 versus 75 ng/L), and LA volume index (43.8 versus 36.2 mL/m²) compared with patients with noncardiac dyspnea (n=34; rest, 8 mm Hg; stress, 18 mm Hg; P≤0.001 for all). Seven patients were excluded because of the presence of non-HFpEF cardiac disease causing dyspnea on imaging. There were no differences in RT-CMR left ventricular total and early diastolic filling at rest and during exercise stress (P≥0.164) between patients with HFpEF and noncardiac dyspnea. RT-CMR revealed significantly impaired LA total and early (P<0.001) diastolic emptying in patients with HFpEF during exercise stress. RT-CMR exercise stress LA long axis strain was independently associated with HFpEF (adjusted odds ratio, 0.657 [95% CI, 0.516-0.838]; P=0.001) after adjustment for clinical and imaging measures and emerged as the best predictor for HFpEF (area under the curve at rest 0.82 versus exercise stress 0.93; P=0.029).

CONCLUSIONS

RT-CMR allows highly accurate identification of HFpEF during physiologic exercise and qualifies as a suitable noninvasive diagnostic alternative. These results will need to be confirmed in multicenter prospective research studies to establish widespread routine clinical use. Registration: URL: https://www.clinicaltrials.gov; Unique identifier: NCT03260621. URL: https://dzhk.de/; Unique identifier: DZHK-17.

CG-SENSE revisited: Results from the first ISMRM reproducibility challenge

PURPOSE

The aim of this work is to shed light on the issue of reproducibility in MR image reconstruction in the context of a challenge. Participants had to recreate the results of "Advances in sensitivity encoding with arbitrary k-space trajectories" by Pruessmann et al. METHODS: The task of the challenge was to reconstruct radially acquired multicoil k-space data (brain/heart) following the method in the original paper, reproducing its key figures. Results were compared to consolidated reference implementations created after the challenge, accounting for the two most common programming languages used in the submissions (Matlab/Python).

RESULTS

Visually, differences between submissions were small. Pixel-wise differences originated from image orientation, assumed field-of-view, or resolution. The reference implementations were in good agreement, both visually and in terms of image similarity metrics.

DISCUSSION AND CONCLUSION

While the description level of the published algorithm enabled participants to reproduce CG-SENSE in general, details of the implementation varied, for example, density compensation or Tikhonov regularization. Implicit assumptions about the data lead to further differences, emphasizing the importance of sufficient metadata accompanying open datasets. Defining reproducibility quantitatively turned out to be nontrivial for this image reconstruction challenge, in the absence of ground-truth results. Typical similarity measures like NMSE of SSIM were misled by image intensity scaling and outlier pixels. Thus, to facilitate reproducibility, researchers are encouraged to publish code and data alongside the original paper. Future methodological papers on MR image reconstruction might benefit from the consolidated reference implementations of CG-SENSE presented here, as a benchmark for methods comparison.

Real-time cardiac magnetic resonance tissue characterisation for fibrosis assessment in aortic stenosis

Background

Myocardial fibrosis is a major determinant of outcome in aortic stenosis (AS). Novel fast real-time (RT) cardiac magnetic resonance (CMR) mapping techniques allow comprehensive quantification of fibrosis but have not yet been adequately validated against standard techniques and histology.

Methods

Patients with severe AS underwent CMR before (n=110) and left ventricular (LV) endomyocardial biopsy (n=46) at transcatheter aortic valve replacement (TAVR). Midventricular short axis native, post-contrast T1 and extracellular volume fraction (ECV) maps were generated using commercially available 5(3)3 MOLLI and RT single-shot inversion recovery fast low-angle shot (FLASH) with radial undersampling. ECV and LV mass were used to calculate LV matrix volumes. Variability and agreements were assessed between RT, MOLLI and histology using intraclass correlation coefficients, coefficients of variation and Bland Altman analyses.

Results

RT and MOLLI derived ECV were similar for myocardium (26.2 vs. 26.5, p=0.073) and inter-ventricular septum (26.2 vs. 26.5, p=0.216). MOLLI native T1 time was in median 20 ms longer compared to RT (p&lt;0.001). Agreement between RT and MOLLI was best for ECV (ICC &gt;0.91), excellent for post-contrast T1 times (ICC &gt;0.81) and good for native T1 times (ICC &gt;0.62). Diffuse collagen volume fraction by biopsies was in median 7.8%. ECV (RT r=0.345, p=0.039; MOLLI r=0.40, p=0.010) and LV matrix volumes (RT r=0.45, p=0.005; MOLLI r=0.43, p=0.007) were the only parameters associated with histology.

Conclusions

RT mapping offers precise T1 and ECV assessments with similar agreement with histology as compared to conventional MOLLI techniques. Single-shot real time techniques may be advantageous in sicker patients prone to dyspnoea or arrhythmia.

Funding Acknowledgement

Type of funding source: Foundation. Main funding source(s): German Research Foundation

Assessment of esophageal motility disorders by real-time MRI

PURPOSE

To investigate imaging findings of esophageal motility disorders on dynamic real-time.

MATERIAL AND METHODS

102 patients with GERD-like symptoms were included in this retrospective study between 2015-2018. Dynamic real-time MRI visualized the transit of a 10 mL pineapple juice bolus through the esophagus and EGJ with a temporal resolution of 40 ms. Dynamic and anatomic parameters were measured by consensus reading. Imaging findings were compared to HRM utilizing the Chicago classification of esophageal motility disorders, v3.0.

RESULTS

All 102 patients completed real-time MRI in a median examination time of 15 min. On HRM, 14 patients presented with disorders with EGJ outlet obstruction (EGJOO) (13.7 %), 7 patients with major disorders of peristalsis (6.9 %), and 32 patients with minor disorders of peristalsis (31.4 %). HRM was normal in 49 patients (48.0 %). Incomplete bolus clearance was significantly more frequent in patients with esophageal motility disorders on HRM than in patients with normal HRM (p = 0.0002). In patients with motility disorders with EGJOO and major disorders of peristalsis, the esophageal diameter tended to be wider (23.6 ± 8.0 vs. 21.2 ± 3.5 mm, p = 0.089) and the sphincter length longer (19.7 ± 7.3 vs. 16.7 ± 3.0 mm, p = 0.091) compared to patients with normal HRM. 3/7 patients with achalasia type II were correctly identified by real-time MRI and one further achalasia type II patient was diagnosed with a motility disorder on MRI films. The other 3/7 patients presented no specific imaging features.

CONCLUSION

Real-time MRI is an auxiliary diagnostic tool for the assessment of swallowing events. Imaging parameters may assist in the detection of esophageal motility disorders.

Cardiac and Respiratory Self-Gating in Radial MRI Using an Adapted Singular Spectrum Analysis (SSA-FARY)

Cardiac Magnetic Resonance Imaging (MRI) is time-consuming and error-prone. To ease the patient's burden and to increase the efficiency and robustness of cardiac exams, interest in methods based on continuous steady-state acquisition and self-gating has been growing in recent years. Self-gating methods extract the cardiac and respiratory signals from the measurement data and then retrospectively sort the data into cardiac and respiratory phases. Repeated breathholds and synchronization with the heart beat using some external device as required in conventional MRI are then not necessary. In this work, we introduce a novel self-gating method for radially acquired data based on a dimensionality reduction technique for time-series analysis (SSA-FARY). Building on Singular Spectrum Analysis, a zero-padded, time-delayed embedding of the auto-calibration data is analyzed using Principle Component Analysis. We demonstrate the basic functionality of SSA-FARY using numerical simulations and apply it to in-vivo cardiac radial single-slice bSSFP and Simultaneous Multi-Slice radiofrequency-spoiled gradient-echo measurements, as well as to Stack-of-Stars bSSFP measurements. SSA-FARY reliably detects the cardiac and respiratory motion and separates it from noise. We utilize the generated signals for high-dimensional image reconstruction using parallel imaging and compressed sensing with in-plane wavelet and (spatio-)temporal total-variation regularization.

Real-time cardiovascular magnetic resonance T1 and extracellular volume fraction mapping for tissue characterisation in aortic stenosis

BACKGROUND

Myocardial fibrosis is a major determinant of outcome in aortic stenosis (AS). Novel fast real-time (RT) cardiovascular magnetic resonance (CMR) mapping techniques allow comprehensive quantification of fibrosis but have not yet been compared against standard techniques and histology.

METHODS

Patients with severe AS underwent CMR before (n = 110) and left ventricular (LV) endomyocardial biopsy (n = 46) at transcatheter aortic valve replacement (TAVR). Midventricular short axis (SAX) native, post-contrast T1 and extracellular volume fraction (ECV) maps were generated using commercially available modified Look-Locker Inversion recovery (MOLLI) (native: 5(3)3, post-contrast: 4(1)3(1)2) and RT single-shot inversion recovery Fast Low-Angle Shot (FLASH) with radial undersampling. Focal late gadolinium enhancement was excluded from T1 and ECV regions of interest. ECV and LV mass were used to calculate LV matrix volumes. Variability and agreements were assessed between RT, MOLLI and histology using intraclass correlation coefficients, coefficients of variation and Bland Altman analyses.

RESULTS

RT and MOLLI derived ECV were similar for midventricular SAX slice coverage (26.2 vs. 26.5, p = 0.073) and septal region of interest (26.2 vs. 26.5, p = 0.216). MOLLI native T1 time was in median 20 ms longer compared to RT (p < 0.001). Agreement between RT and MOLLI was best for ECV (ICC > 0.91), excellent for post-contrast T1 times (ICC > 0.81) and good for native T1 times (ICC > 0.62). Diffuse collagen volume fraction by biopsies was in median 7.8%. ECV (RT r = 0.345, p = 0.039; MOLLI r = 0.40, p = 0.010) and LV matrix volumes (RT r = 0.45, p = 0.005; MOLLI r = 0.43, p = 0.007) were the only parameters associated with histology.

CONCLUSIONS

RT mapping offers fast and sufficient ECV and LV matrix volume calculation in AS patients. ECV and LV matrix volume represent robust and universally comparable parameters with associations to histologically assessed fibrosis and may emerge as potential targets for clinical decision making.

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.

Joint T1 and T2 Mapping With Tiny Dictionaries and Subspace-Constrained Reconstruction

A novel method is developed that adaptively generates tiny dictionaries for joint T1-T2 mapping in magnetic resonance imaging. This work breaks the bond between dictionary size and representation accuracy (i) by approximating the Bloch-response manifold by piece-wise linear functions and (ii) by adaptively refining the sampling grid depending on the locally-linear approximation error. Data acquisition is accomplished with use of an 2D radially sampled Inversion-Recovery Hybrid-State Free Precession sequence. Adaptive dictionaries are generated with different error tolerances and compared to a heuristically designed dictionary. Based on simulation results, tiny dictionaries were used for T1-T2 mapping in phantom and in vivo studies. Reconstruction and parameter mapping were performed entirely in subspace. All experiments demonstrated excellent agreement between the proposed mapping technique and template matching using heuristic dictionaries. Adaptive dictionaries in combination with manifold projection allow to reduce the necessary dictionary sizes by one to two orders of magnitude.

Real-time MRI for dynamic assessment of gastroesophageal reflux disease: Comparison to pH-metry and impedance

PURPOSE

To evaluate the diagnostic potential of real-time MRI for dynamic assessment of gastroesophageal reflux in patients with GERD (gastroesophageal reflux disease)-like symptoms compared to pH-metry and impedance.

METHODS

Patients who underwent real-time MRI and pH-metry between 2015-2018 were included in this retrospective study. Real-time MRI at 3 T was achieved by undersampled radial FLASH acquisitions with iterative image reconstruction by NLINV. Real-time MRI visualized transit of pineapple juice through the gastroesophageal junction and during Valsalva maneuver. MRI results were compared to 24 h pH-metry to assess acidic reflux (following Lyon Consensus guidelines) and to impedance to assess non-acidic reflux. A standard 2 × 2 table was chosen to calculate diagnostic performance.

RESULTS

91/93 eligible patients fulfilled inclusion criteria (male n = 49; female n = 42; median age 55 y). All MRI studies were successfully completed without adverse events at a mean examination time of 15 min. On real-time MRI, reflux was evident in 60 patients (66 %). pH-metry revealed reflux in 41 patients (45 %), and impedance in 54 patients (59 %). Compared to pH-metry and impedance, real-time MRI sensitivity was 0.78 (95 % CI: 0.66-0.87), specificity 0.67 (95 % CI: 0.45-0.84) and PPV 0.87 (95 % CI: 0.75-0.94).

CONCLUSION

Real-time MRI is an imaging method for assessment of gastroesophageal reflux in patients with GERD-like symptoms. Considering its high positive predictive value, real-time MRI can accurately identify patients in which further invasive testing with pH-metry and impedance might be considered.

Model-based myocardial T1 mapping with sparsity constraints using single-shot inversion-recovery radial FLASH cardiovascular magnetic resonance

BACKGROUND

This study develops a model-based myocardial T1 mapping technique with sparsity constraints which employs a single-shot inversion-recovery (IR) radial fast low angle shot (FLASH) cardiovascular magnetic resonance (CMR) acquisition. The method should offer high resolution, accuracy, precision and reproducibility.

METHODS

The proposed reconstruction estimates myocardial parameter maps directly from undersampled k-space which is continuously measured by IR radial FLASH with a 4 s breathhold and retrospectively sorted based on a cardiac trigger signal. Joint sparsity constraints are imposed on the parameter maps to further improve T1 precision. Validations involved studies of an experimental phantom and 8 healthy adult subjects.

RESULTS

In comparison to an IR spin-echo reference method, phantom experiments with T1 values ranging from 300 to 1500 ms revealed good accuracy and precision at simulated heart rates between 40 and 100 bpm. In vivo T1 maps achieved better precision and qualitatively better preservation of image features for the proposed method than a real-time CMR approach followed by pixelwise fitting. Apart from good inter-observer reproducibility (0.6% of the mean), in vivo results confirmed good intra-subject reproducibility (1.05% of the mean for intra-scan and 1.17, 1.51% of the means for the two inter-scans, respectively) of the proposed method.

CONCLUSION

Model-based reconstructions with sparsity constraints allow for single-shot myocardial T1 maps with high spatial resolution, accuracy, precision and reproducibility within a 4 s breathhold. Clinical trials are warranted.

Dynamic water/fat separation and B 0 inhomogeneity mapping-joint estimation using undersampled triple-echo multi-spoke radial FLASH

PURPOSE

To achieve dynamic water/fat separation and B 0 field inhomogeneity mapping via model-based reconstructions of undersampled triple-echo multi-spoke radial FLASH acquisitions.

METHODS

This work introduces an undersampled triple-echo multi-spoke radial FLASH sequence, which uses (i) complementary radial spokes per echo train for faster spatial encoding, (ii) asymmetric echoes for flexible and nonuniform echo spacing, and (iii) a golden angle increment across frames for optimal k-space coverage. Joint estimation of water, fat, B 0 inhomogeneity, and coil sensitivity maps from undersampled triple-echo data poses a nonlinear and non-convex inverse problem which is solved by a model-based reconstruction with suitable regularization. The developed methods are validated using phantom experiments with different degrees of undersampling. Real-time MRI studies of the knee, liver, and heart are conducted without prospective gating or retrospective data sorting at temporal resolutions of 70, 158, and 40 ms, respectively.

RESULTS

Up to 18-fold undersampling is achieved in this work. Even in the presence of rapid physiological motion, large B 0 field inhomogeneities, and phase wrapping, the model-based reconstruction yields reliably separated water/fat maps in conjunction with spatially smooth inhomogeneity maps.

CONCLUSIONS

The combination of a triple-echo acquisition and joint reconstruction technique provides a practical solution to time-resolved and motion robust water/fat separation at high spatial and temporal resolution.

Frequency-modulated SSFP with radial sampling and subspace reconstruction: A time-efficient alternative to phase-cycled bSSFP

PURPOSE

A novel subspace-based reconstruction method for frequency-modulated balanced steady-state free precession (fmSSFP) MRI is presented. In this work, suitable data acquisition schemes, subspace sizes, and efficiencies for banding removal are investigated.

THEORY AND METHODS

By combining a fmSSFP MRI sequence with a 3D stack-of-stars trajectory, scan efficiency is maximized as spectral information is obtained without intermediate preparation phases. A memory-efficient reconstruction routine is implemented by introducing the low-frequency Fourier transform as a subspace which allows for the formulation of a convex reconstruction problem. The removal of banding artifacts is investigated by comparing the proposed acquisition and reconstruction technique to phase-cycled bSSFP MRI. Aliasing properties of different undersampling schemes are analyzed and water/fat separation is demonstrated by reweighting the reconstructed subspace coefficients to generate virtual spectral responses in a post-processing step.

RESULTS

A simple root-of-sum-of-squares combination of the reconstructed subspace coefficients yields high-SNR images with the characteristic bSSFP contrast but without banding artifacts. Compared to Golden-Angle trajectories, turn-based sampling schemes were superior in minimizing aliasing across reconstructed subspace coefficients. Water/fat separated images of the human knee were obtained by reweighting subspace coefficients.

CONCLUSIONS

The novel subspace-based fmSSFP MRI technique emerges as a time-efficient alternative to phase-cycled bSFFP. The method does not need intermediate preparation phases, offers high SNR and avoids banding artifacts. Reweighting of the reconstructed subspace coefficients allows for generating virtual spectral responses with applications to water/fat separation.

Model-based T1 mapping with sparsity constraints using single-shot inversion-recovery radial FLASH

PURPOSE

To develop a model-based reconstruction technique for single-shot T₁ mapping with high spatial resolution, accuracy, and precision using an inversion-recovery (IR) fast low-angle shot (FLASH) acquisition with radial encoding.

METHODS

The proposed model-based reconstruction jointly estimates all model parameters, that is, the equilibrium magnetization, steady-state magnetization, 1/ T1*, and all coil sensitivities from the data of a single-shot IR FLASH acquisition with a small golden-angle radial trajectory. Joint sparsity constraints on the parameter maps are exploited to improve the performance of the iteratively regularized Gauss-Newton method chosen for solving the nonlinear inverse problem. Validations include both a numerical and experimental T₁ phantom, as well as in vivo studies of the human brain and liver at 3 T.

RESULTS

In comparison to previous reconstruction methods for single-shot T₁ mapping, which are based on real-time MRI with pixel-wise fitting and a model-based approach with a predetermination of coil sensitivities, the proposed method presents with improved robustness against phase errors and numerical precision in both phantom and in vivo studies.

CONCLUSION

The comprehensive model-based reconstruction with L1 regularization offers rapid and robust T₁ mapping with high accuracy and precision. The method warrants accelerated computing and online implementation for extended clinical trials. Magn Reson Med 79:730-740, 2018. © 2017 International Society for Magnetic Resonance in Medicine.

Accelerated whole-heart MR angiography using a variable-density poisson-disc undersampling pattern and compressed sensing reconstruction

PURPOSE

To accelerate whole-heart three-dimension MR angiography (MRA) by using a variable-density Poisson-disc undersampling pattern and a compressed sensing (CS) reconstruction algorithm, and compare the results with sensitivity encoding (SENSE).

METHODS

For whole-heart MRA, a prospective variable-density Poisson-disc k-space undersampling pattern was developed in which 1-2% of central part of k-space was fully sampled, and sampling in the remainder decreased exponentially toward the periphery. The undersampled data were then estimated using CS reconstruction. In patients, images using this sequence with an undersampling rate of ≈6 were compared with those using a SENSE rate of 2 (n = 15) and a SENSE rate of 6 (n = 13).

RESULTS

Compared with SENSE rate 2, CS rate 6 images had similar objective border sharpness, significantly lower subjective image quality scores at all four locations (all P < 0.01), and shorter scan times (P < 0.05). Compared with SENSE rate 6, CS rate 6 had similar objective border sharpness at all four locations, significantly better subjective image quality scores at three of four locations (all P < 0.01), and similar scan times (P = 0.24).

CONCLUSION

Compared with SENSE with a comparable acceleration rate, a variable-density Poisson-disc undersampling pattern and CS reconstruction achieved better subjective image quality and similar border sharpness. Magn Reson Med 79:761-769, 2018. © 2017 International Society for Magnetic Resonance in Medicine.

Targeted endomyocardial biopsy guided by real-time cardiovascular magnetic resonance

BACKGROUND

Endomyocardial biopsies (EMB) are an important diagnostic tool for myocarditis and other infiltrative cardiac diseases. Routinely, biopsies are obtained under fluoroscopic guidance with a substantial radiation burden. Despite procedural success, there is a large sampling error caused by missing the affected myocardium. Therefore, multiple (>6) biopsies are taken in the clinical setting. In cardiovascular magnetic resonance (CMR), late gadolinium enhancement (LGE) depicts areas of affected myocardium in myocarditis or in other infiltrative cardiomyopathies. Thus, targeted biopsy under real-time CMR image guidance might reduce the problem of sampling error.

METHODS

Seven minipigs of the Goettingen strain underwent radiofrequency ablation in the left ventricle. At least two focal lesions were induced on the lateral wall in five and the apex in two animals. Each ablation lesion was created by two consecutive 30 sec ablations (max. 30 W, temperature 60-64 °C). Biopsies were taken immediately after lesion induction using a commercially available 7 F conventional bioptome under fluoroscopic guidance at the ablation site. Afterwards the animals underwent CMR and lesion visualization by LGE at 3T. The lesions were then targeted and biopsied under CMR-guidance using a MR-conditional bioptome guided by a steerable catheter. Interactive real-time (RT) visualization of the intervention on an in-room monitor was based on radial FLASH with nonlinear inverse reconstruction (NLINV) at a temporal resolution of 42 ms. All samples underwent a standard histological evaluation.

RESULTS

Radiofrequency ablation was successful in all animals. Fluoroscopy-guided biopsies were performed with a success rate of 6/6 minipigs - resulting in a nonlethal pericardial effusion in one animal. Visualization of radiofrequency lesions by CMR was successful in 7/7 minipig, i.e. at least one lesion was clearly visible. Localization and tracking of the catheters and the bioptome using interactive control of the imaging plane was achieved in 6/6 MP; however in the animal with a large pericardial effusion after EMB under fluoroscopy no further EMB was attempted for safety reasons. Biopsies under interactive RT-CMR guidance were successfully performed in 5/6 animals, in one animal the bioptome reached the lesion, however the forceps did not cut out a sample. Specimens obtained under CMR guidance contained part of the lesion in 6/15 (40%) myocardial specimens and in 4/5 (80%) animals in which samples were achieved. Conventional biopsies revealed ablation lesions in 4/17 (23.5%) specimens in 3/6 minipigs (50%).

CONCLUSION

Focal lesions induced by radiofrequency ablation in a minipig model are a useful tool for CMR-guided biopsy studies. In contrast to fluoroscopy, CMR provides excellent visualization of lesions. Interactive real-time CMR allows excellent passive tracking of the instruments and EMB provides significantly superior sampling accuracy compared to fluoroscopy-guided biopsies. Nonetheless, further improvements of MR-compatible bioptomes and guiding catheters are essential before applying this method in a clinical setting.

T2 shuffling: Sharp, multicontrast, volumetric fast spin-echo imaging

PURPOSE

A new acquisition and reconstruction method called T2 Shuffling is presented for volumetric fast spin-echo (three-dimensional [3D] FSE) imaging. T2 Shuffling reduces blurring and recovers many images at multiple T2 contrasts from a single acquisition at clinically feasible scan times (6-7 min).

THEORY AND METHODS

The parallel imaging forward model is modified to account for temporal signal relaxation during the echo train. Scan efficiency is improved by acquiring data during the transient signal decay and by increasing echo train lengths without loss in signal-to-noise ratio (SNR). By (1) randomly shuffling the phase encode view ordering, (2) constraining the temporal signal evolution to a low-dimensional subspace, and (3) promoting spatio-temporal correlations through locally low rank regularization, a time series of virtual echo time images is recovered from a single scan. A convex formulation is presented that is robust to partial voluming and radiofrequency field inhomogeneity.

RESULTS

Retrospective undersampling and in vivo scans confirm the increase in sharpness afforded by T2 Shuffling. Multiple image contrasts are recovered and used to highlight pathology in pediatric patients. A proof-of-principle method is integrated into a clinical musculoskeletal imaging workflow.

CONCLUSION

The proposed T2 Shuffling method improves the diagnostic utility of 3D FSE by reducing blurring and producing multiple image contrasts from a single scan. Magn Reson Med 77:180-195, 2017. © 2016 Wiley Periodicals, Inc.

Fast comprehensive single-sequence four-dimensional pediatric knee MRI with T2 shuffling

PURPOSE

To develop and clinically evaluate a pediatric knee magnetic resonance imaging (MRI) technique based on volumetric fast spin-echo (3DFSE) and compare its diagnostic performance, image quality, and imaging time to that of a conventional 2D protocol.

MATERIALS AND METHODS

A 3DFSE sequence was modified and combined with a compressed sensing-based reconstruction resolving multiple image contrasts, a technique termed T₂ Shuffling (T₂ Sh). With Institutional Review Board (IRB) approval, 28 consecutive children referred for 3T knee MRI prospectively underwent a standard clinical knee protocol followed by T₂ Sh. T₂ Sh performance was assessed by two readers blinded to diagnostic reports. Interpretive discrepancies were resolved by medical record chart review and consensus between the readers and an orthopedic surgeon. Image quality was evaluated by rating anatomic delineation, with 95% confidence interval. A Wilcoxon rank-sum test assessed the null hypothesis that T₂ Sh structure delineation compared to conventional 2D is unchanged. Intraclass correlation coefficients were calculated for interobserver agreement. Imaging time of the conventional protocol and T₂ Sh was compared.

RESULTS

There was 81% and 87% concordance between T₂ Sh reports and diagnostic reports, respectively, for each reader. Upon consensus review, T₂ Sh had 93% sensitivity and 100% specificity compared to clinical reports for detection of clinically relevant findings. The 95% confidence interval of diagnostic or better rating was 95-100%, with 34-80% interobserver agreement. There was no significant difference in structure delineation between T₂ Sh and 2D, except for the retinaculum (P < 0.05), where 2D was preferred. Typical imaging time for T₂ Sh and the conventional exam was 7 and 13 minutes, respectively.

CONCLUSION

A single-sequence pediatric knee exam is feasible with T₂ Sh, providing multiplanar, reformattable 4D images.

LEVEL OF EVIDENCE

2 J. MAGN. RESON. IMAGING 2017;45:1700-1711.

Estimating absolute-phase maps using ESPIRiT and virtual conjugate coils

PURPOSE

To develop an ESPIRiT-based method to estimate coil sensitivities with image phase as a building block for efficient and robust image reconstruction with phase constraints.

THEORY AND METHODS

ESPIRiT is a new framework for calibration of the coil sensitivities and reconstruction in parallel magnetic resonance imaging. Applying ESPIRiT to a combined set of physical and virtual conjugate coils (VCC-ESPIRiT) implicitly exploits conjugate symmetry in k-space similar to VCC-GRAPPA. Based on this method, a new post-processing step is proposed for the explicit computation of coil sensitivities that include the absolute phase of the image. The accuracy of the computed maps is directly validated using a test based on projection onto fully sampled coil images and also indirectly in phase-constrained parallel-imaging reconstructions.

RESULTS

The proposed method can estimate accurate sensitivities which include low-resolution image phase. In case of high-frequency phase variations VCC-ESPIRiT yields an additional set of maps that indicates the existence of a high-frequency phase component. Taking this additional set of maps into account can improve the robustness of phase-constrained parallel imaging.

CONCLUSION

The extended VCC-ESPIRiT is a useful tool for phase-constrained imaging. Magn Reson Med 77:1201-1207, 2017. © 2016 International Society for Magnetic Resonance in Medicine.

Comprehensive motion-compensated highly accelerated 4D flow MRI with ferumoxytol enhancement for pediatric congenital heart disease

PURPOSE

To develop and evaluate motion-compensation and compressed-sensing techniques in 4D flow MRI for anatomical assessment in a comprehensive ferumoxytol-enhanced congenital heart disease (CHD) exam.

MATERIALS AND METHODS

A Cartesian 4D flow sequence was developed to enable intrinsic navigation and two variable-density sampling schemes: VDPoisson and VDRad. Four compressed-sensing methods were developed: A) VDPoisson scan reconstructed using spatial wavelets; B) added temporal total variation to A; C) VDRad scan using the same reconstruction as in B; and D) added motion compensation to C. With Institutional Review Board (IRB) approval and Health Insurance Portability and Accountability Act (HIPAA) compliance, 23 consecutive patients (eight females, mean 6.3 years) referred for ferumoxytol-enhanced CHD 3T MRI were recruited. Images were acquired and reconstructed using methods A-D. Two cardiovascular radiologists independently scored the images on a 5-point scale. These readers performed a paired wall motion and functional assessment between method D and 2D balanced steady-state free precession (bSSFP) CINE for 16 cases.

RESULTS

Method D had higher diagnostic image quality for most anatomical features (mean 3.8-4.8) compared to A (2.0-3.6), B (2.2-3.7), and C (2.9-3.9) with P < 0.05 with good interobserver agreement (κ ≥ 0.49). Method D had similar or better assessment of myocardial borders and cardiac motion compared to 2D bSSFP (P < 0.05, κ ≥ 0.77). All methods had good internal agreement in comparing aortic with pulmonic flow (BA mean < 0.02%, r > 0.85) and compared to method A (BA mean < 0.13%, r > 0.84) with P < 0.01.

CONCLUSION

Flow, functional, and anatomical assessment in CHD with ferumoxytol-enhanced 4D flow is feasible and can be significantly improved using motion compensation and compressed sensing. J. Magn. Reson. Imaging 2016;43:1355-1368.

ESPIRiT--an eigenvalue approach to autocalibrating parallel MRI: where SENSE meets GRAPPA

PURPOSE

Parallel imaging allows the reconstruction of images from undersampled multicoil data. The two main approaches are: SENSE, which explicitly uses coil sensitivities, and GRAPPA, which makes use of learned correlations in k-space. The purpose of this work is to clarify their relationship and to develop and evaluate an improved algorithm.

THEORY AND METHODS

A theoretical analysis shows: (1) The correlations in k-space are encoded in the null space of a calibration matrix. (2) Both approaches restrict the solution to a subspace spanned by the sensitivities. (3) The sensitivities appear as the main eigenvector of a reconstruction operator computed from the null space. The basic assumptions and the quality of the sensitivity maps are evaluated in experimental examples. The appearance of additional eigenvectors motivates an extended SENSE reconstruction with multiple maps, which is compared to existing methods.

RESULTS

The existence of a null space and the high quality of the extracted sensitivities are confirmed. The extended reconstruction combines all advantages of SENSE with robustness to certain errors similar to GRAPPA.

CONCLUSION

In this article the gap between both approaches is finally bridged. A new autocalibration technique combines the benefits of both.

Parallel Magnetic Resonance Imaging as Approximation in a Reproducing Kernel Hilbert Space

In Magnetic Resonance Imaging (MRI) data samples are collected in the spatial frequency domain (k-space), typically by time-consuming line-by-line scanning on a Cartesian grid. Scans can be accelerated by simultaneous acquisition of data using multiple receivers (parallel imaging), and by using more efficient non-Cartesian sampling schemes. To understand and design k-space sampling patterns, a theoretical framework is needed to analyze how well arbitrary sampling patterns reconstruct unsampled k-space using receive coil information. As shown here, reconstruction from samples at arbitrary locations can be understood as approximation of vector-valued functions from the acquired samples and formulated using a Reproducing Kernel Hilbert Space (RKHS) with a matrix-valued kernel defined by the spatial sensitivities of the receive coils. This establishes a formal connection between approximation theory and parallel imaging. Theoretical tools from approximation theory can then be used to understand reconstruction in k-space and to extend the analysis of the effects of samples selection beyond the traditional image-domain g-factor noise analysis to both noise amplification and approximation errors in k-space. This is demonstrated with numerical examples.

Fast pediatric 3D free-breathing abdominal dynamic contrast enhanced MRI with high spatiotemporal resolution

PURPOSE

To develop a method for fast pediatric 3D free-breathing abdominal dynamic contrast enhanced (DCE) magnetic resonance imaging (MRI) and investigate its clinical feasibility.

MATERIALS AND METHODS

A combined locally low rank parallel imaging method with soft gating is proposed for free-breathing DCE MRI acquisition. With Institutional Review Board (IRB) approval and informed consent/assent, 23 consecutive pediatric patients were recruited for this study. Free-breathing DCE MRI with ∼1 mm(3) spatial resolution and a 6.5-sec frame rate was acquired on a 3T scanner. Undersampled data were reconstructed with a compressed sensing method without motion correction (FB-CS) and the proposed method (FB-LR). A follow-up respiratory-triggered acquisition (RT-CS) was performed as a reference standard. The reconstructed images were evaluated independently by two radiologists. Wilcoxon tests were performed to test the hypothesis that there was no significant difference between different reconstructions. Quantitative evaluation of contrast dynamics was also performed.

RESULTS

The mean score of overall image quality of FB-LR was 4.0 on a 5-point scale, significantly better (P < 0.05) than FB-CS reconstruction (mean score 2.9), and similar to RT-CS (mean score 4.1). FB-LR also matched the temporal fidelity of contrast dynamics with a root mean square error less than 5%.

CONCLUSION

Fast 3D free-breathing DCE MRI with high scan efficiency and image quality similar to respiratory-triggered acquisition is feasible in a pediatric clinical setting.

Fast T2 mapping with improved accuracy using undersampled spin-echo MRI and model-based reconstructions with a generating function

A model-based reconstruction technique for accelerated T2 mapping with improved accuracy is proposed using undersampled Cartesian spin-echo magnetic resonance imaging (MRI) data. The technique employs an advanced signal model for T2 relaxation that accounts for contributions from indirect echoes in a train of multiple spin echoes. An iterative solution of the nonlinear inverse reconstruction problem directly estimates spin-density and T2 maps from undersampled raw data. The algorithm is validated for simulated data as well as phantom and human brain MRI at 3T. The performance of the advanced model is compared to conventional pixel-based fitting of echo-time images from fully sampled data. The proposed method yields more accurate T2 values than the mono-exponential model and allows for retrospective undersampling factors of at least 6.Although limitations are observed for very long T2 relaxation times, respective reconstruction problems may be overcome by a gradient dampening approach. The analytical gradient of the utilized cost function is included as appendix. The source code is made available to the community.

Free-breathing pediatric MRI with nonrigid motion correction and acceleration

PURPOSE

To develop and assess motion correction techniques for high-resolution pediatric abdominal volumetric magnetic resonance images acquired free-breathing with high scan efficiency.

MATERIALS AND METHODS

First, variable-density sampling and radial-like phase-encode ordering were incorporated into the 3D Cartesian acquisition. Second, intrinsic multichannel butterfly navigators were used to measure respiratory motion. Lastly, these estimates are applied for both motion-weighted data-consistency in a compressed sensing and parallel imaging reconstruction, and for nonrigid motion correction using a localized autofocusing framework. With Institutional Review Board approval and informed consent/assent, studies were performed on 22 consecutive pediatric patients. Two radiologists independently scored the images for overall image quality, degree of motion artifacts, and sharpness of hepatic vessels and the diaphragm. The results were assessed using paired Wilcoxon test and weighted kappa coefficient for interobserver agreements.

RESULTS

The complete procedure yielded significantly better overall image quality (mean score of 4.7 out of 5) when compared to using no correction (mean score of 3.4, P < 0.05) and to using motion-weighted accelerated imaging (mean score of 3.9, P < 0.05). With an average scan time of 28 seconds, the proposed method resulted in comparable image quality to conventional prospective respiratory-triggered acquisitions with an average scan time of 91 seconds (mean score of 4.5).

CONCLUSION

With the proposed methods, diagnosable high-resolution abdominal volumetric scans can be obtained from free-breathing data acquisitions.

Chemical shift separation with controlled aliasing for hyperpolarized (13) C metabolic imaging

PURPOSE

A chemical shift separation technique for hyperpolarized (13) C metabolic imaging with high spatial and temporal resolution was developed. Specifically, a fast three-dimensional pulse sequence and a reconstruction method were implemented to acquire signals from multiple (13) C species simultaneously with subsequent separation into individual images.

THEORY AND METHODS

A stack of flyback echo-planar imaging readouts and a set of multiband excitation radiofrequency pulses were designed to spatially modulate aliasing patterns of the acquired metabolite images, which translated the chemical shift separation problem into parallel imaging reconstruction problem. An eight-channel coil array was used for data acquisition and a parallel imaging method based on nonlinear inversion was developed to separate the aliased images.

RESULTS

Simultaneous acquisitions of pyruvate and lactate in a phantom study and in vivo rat experiments were performed. The results demonstrated successful separation of the metabolite distributions into individual images having high spatial resolution.

CONCLUSION

This method demonstrated the ability to provide accelerated metabolite imaging in hyperpolarized (13) C MR using multichannel coils, tailored readout, and specialized RF pulses.

Robust 4D flow denoising using divergence-free wavelet transform

PURPOSE

To investigate four-dimensional flow denoising using the divergence-free wavelet (DFW) transform and compare its performance with existing techniques.

THEORY AND METHODS

DFW is a vector-wavelet that provides a sparse representation of flow in a generally divergence-free field and can be used to enforce "soft" divergence-free conditions when discretization and partial voluming result in numerical nondivergence-free components. Efficient denoising is achieved by appropriate shrinkage of divergence-free wavelet and nondivergence-free coefficients. SureShrink and cycle spinning are investigated to further improve denoising performance.

RESULTS

DFW denoising was compared with existing methods on simulated and phantom data and was shown to yield better noise reduction overall while being robust to segmentation errors. The processing was applied to in vivo data and was demonstrated to improve visualization while preserving quantifications of flow data.

CONCLUSION

DFW denoising of four-dimensional flow data was shown to reduce noise levels in flow data both quantitatively and visually.