Unsupervised deep learning with convolutional neural networks for static parallel transmit design: A retrospective study.

PURPOSE: To mitigate B 1 + $$ {B}_1^{+} $$ inhomogeneity at 7T for multi-channel transmit arrays using unsupervised deep learning with convolutional neural networks (CNNs). METHODS: Deep learning parallel transmit (pTx) pulse design has received attention, but such methods have relied on supervised training and did not use CNNs for multi-channel B 1 + $$ {B}_1^{+} $$ maps. In this work, we introduce an alternative approach that facilitates the use of CNNs with multi-channel B 1 + $$ {B}_1^{+} $$ maps while performing unsupervised training. The multi-channel B 1 + $$ {B}_1^{+} $$ maps are concatenated along the spatial dimension to enable shift-equivariant processing amenable to CNNs. Training is performed in an unsupervised manner using a physics-driven loss function that minimizes the discrepancy of the Bloch simulation with the target magnetization, which eliminates the calculation of reference transmit RF weights. The training database comprises 3824 2D sagittal, multi-channel B 1 + $$ {B}_1^{+} $$ maps of the healthy human brain from 143 subjects. B 1 + $$ {B}_1^{+} $$ data were acquired at 7T using an 8Tx/32Rx head coil. The proposed method is compared to the unregularized magnitude least-squares (MLS) solution for the target magnetization in static pTx design. RESULTS: The proposed method outperformed the unregularized MLS solution for RMS error and coefficient-of-variation and had comparable energy consumption. Additionally, the proposed method did not show local phase singularities leading to distinct holes in the resulting magnetization unlike the unregularized MLS solution. CONCLUSION: Proposed unsupervised deep learning with CNNs performs better than unregularized MLS in static pTx for speed and robustness.

Large-scale 3D non-Cartesian coronary MRI reconstruction using distributed memory-efficient physics-guided deep learning with limited training data.

OBJECT: To enable high-quality physics-guided deep learning (PG-DL) reconstruction of large-scale 3D non-Cartesian coronary MRI by overcoming challenges of hardware limitations and limited training data availability. MATERIALS AND METHODS: While PG-DL has emerged as a powerful image reconstruction method, its application to large-scale 3D non-Cartesian MRI is hindered by hardware limitations and limited availability of training data. We combine several recent advances in deep learning and MRI reconstruction to tackle the former challenge, and we further propose a 2.5D reconstruction using 2D convolutional neural networks, which treat 3D volumes as batches of 2D images to train the network with a limited amount of training data. Both 3D and 2.5D variants of the PG-DL networks were compared to conventional methods for high-resolution 3D kooshball coronary MRI. RESULTS: Proposed PG-DL reconstructions of 3D non-Cartesian coronary MRI with 3D and 2.5D processing outperformed all conventional methods both quantitatively and qualitatively in terms of image assessment by an experienced cardiologist. The 2.5D variant further improved vessel sharpness compared to 3D processing, and scored higher in terms of qualitative image quality. DISCUSSION: PG-DL reconstruction of large-scale 3D non-Cartesian MRI without compromising image size or network complexity is achieved, and the proposed 2.5D processing enables high-quality reconstruction with limited training data.

Signal intensity informed multi-coil encoding operator for physics-guided deep learning reconstruction of highly accelerated myocardial perfusion CMR.

PURPOSE: To develop a physics-guided deep learning (PG-DL) reconstruction strategy based on a signal intensity informed multi-coil (SIIM) encoding operator for highly-accelerated simultaneous multislice (SMS) myocardial perfusion cardiac MRI (CMR). METHODS: First-pass perfusion CMR acquires highly-accelerated images with dynamically varying signal intensity/SNR following the administration of a gadolinium-based contrast agent. Thus, using PG-DL reconstruction with a conventional multi-coil encoding operator leads to analogous signal intensity variations across different time-frames at the network output, creating difficulties in generalization for varying SNR levels. We propose to use a SIIM encoding operator to capture the signal intensity/SNR variations across time-frames in a reformulated encoding operator. This leads to a more uniform/flat contrast at the output of the PG-DL network, facilitating generalizability across time-frames. PG-DL reconstruction with the proposed SIIM encoding operator is compared to PG-DL with conventional encoding operator, split slice-GRAPPA, locally low-rank (LLR) regularized reconstruction, low-rank plus sparse (L + S) reconstruction, and regularized ROCK-SPIRiT. RESULTS: Results on highly accelerated free-breathing first pass myocardial perfusion CMR at three-fold SMS and four-fold in-plane acceleration show that the proposed method improves upon the reconstruction methods use for comparison. Substantial noise reduction is achieved compared to split slice-GRAPPA, and aliasing artifacts reduction compared to LLR regularized reconstruction, L + S reconstruction and PG-DL with conventional encoding. Furthermore, a qualitative reader study indicated that proposed method outperformed all methods. CONCLUSION: PG-DL reconstruction with the proposed SIIM encoding operator improves generalization across different time-frames /SNRs in highly accelerated perfusion CMR.

Residual RAKI: A hybrid linear and non-linear approach for scan-specific k-space deep learning.

Parallel imaging is the most clinically used acceleration technique for magnetic resonance imaging (MRI) in part due to its easy inclusion into routine acquisitions. In k-space based parallel imaging reconstruction, sub-sampled k-space data are interpolated using linear convolutions. At high acceleration rates these methods have inherent noise amplification and reduced image quality. On the other hand, non-linear deep learning methods provide improved image quality at high acceleration, but the availability of training databases for different scans, as well as their interpretability hinder their adaptation. In this work, we present an extension of Robust Artificial-neural-networks for k-space Interpolation (RAKI), called residual-RAKI (rRAKI), which achieves scan-specific machine learning reconstruction using a hybrid linear and non-linear methodology. In rRAKI, non-linear CNNs are trained jointly with a linear convolution implemented via a skip connection. In effect, the linear part provides a baseline reconstruction, while the non-linear CNN that runs in parallel provides further reduction of artifacts and noise arising from the linear part. The explicit split between the linear and non-linear aspects of the reconstruction also help improve interpretability compared to purely non-linear methods. Experiments were conducted on the publicly available fastMRI datasets, as well as high-resolution anatomical imaging, comparing GRAPPA and its variants, compressed sensing, RAKI, Scan Specific Artifact Reduction in K-space (SPARK) and the proposed rRAKI. Additionally, highly-accelerated simultaneous multi-slice (SMS) functional MRI reconstructions were also performed, where the proposed rRAKI was compred to Read-out SENSE-GRAPPA and RAKI. Our results show that the proposed rRAKI method substantially improves the image quality compared to conventional parallel imaging, and offers sharper images compared to SPARK and ℓ1-SPIRiT. Furthermore, rRAKI shows improved preservation of time-varying dynamics compared to both parallel imaging and RAKI in highly-accelerated SMS fMRI.

Multi-mask self-supervised learning for physics-guided neural networks in highly accelerated magnetic resonance imaging.

Self-supervised learning has shown great promise because of its ability to train deep learning (DL) magnetic resonance imaging (MRI) reconstruction methods without fully sampled data. Current self-supervised learning methods for physics-guided reconstruction networks split acquired undersampled data into two disjoint sets, where one is used for data consistency (DC) in the unrolled network, while the other is used to define the training loss. In this study, we propose an improved self-supervised learning strategy that more efficiently uses the acquired data to train a physics-guided reconstruction network without a database of fully sampled data. The proposed multi-mask self-supervised learning via data undersampling (SSDU) applies a holdout masking operation on the acquired measurements to split them into multiple pairs of disjoint sets for each training sample, while using one of these pairs for DC units and the other for defining loss, thereby more efficiently using the undersampled data. Multi-mask SSDU is applied on fully sampled 3D knee and prospectively undersampled 3D brain MRI datasets, for various acceleration rates and patterns, and compared with the parallel imaging method, CG-SENSE, and single-mask SSDU DL-MRI, as well as supervised DL-MRI when fully sampled data are available. The results on knee MRI show that the proposed multi-mask SSDU outperforms SSDU and performs as well as supervised DL-MRI. A clinical reader study further ranks the multi-mask SSDU higher than supervised DL-MRI in terms of signal-to-noise ratio and aliasing artifacts. Results on brain MRI show that multi-mask SSDU achieves better reconstruction quality compared with SSDU. The reader study demonstrates that multi-mask SSDU at R = 8 significantly improves reconstruction compared with single-mask SSDU at R = 8, as well as CG-SENSE at R = 2.

Improved simultaneous multislice cardiac MRI using readout concatenated k-space SPIRiT (ROCK-SPIRiT).

PURPOSE: To develop and evaluate a simultaneous multislice (SMS) reconstruction technique that provides noise reduction and leakage blocking for highly accelerated cardiac MRI. METHODS: ReadOut Concatenated k-space SPIRiT (ROCK-SPIRiT) uses the concept of readout concatenation in image domain to represent SMS encoding, and performs coil self-consistency as in SPIRiT-type reconstruction in an extended k-space, while allowing regularization for further denoising. The proposed method is implemented with and without regularization, and validated on retrospectively SMS-accelerated cine imaging with three-fold SMS and two-fold in-plane acceleration. ROCK-SPIRiT is compared with two leakage-blocking SMS reconstruction methods: readout-SENSE-GRAPPA and split slice-GRAPPA. Further evaluation and comparisons are performed using prospectively SMS-accelerated cine imaging. RESULTS: Results on retrospectively three-fold SMS and two-fold in-plane accelerated cine imaging show that ROCK-SPIRiT without regularization significantly improves on existing methods in terms of PSNR (readout-SENSE-GRAPPA: 33.5 ± 3.2, split slice-GRAPPA: 34.1 ± 3.8, ROCK-SPIRiT: 35.0 ± 3.3) and SSIM (readout-SENSE-GRAPPA: 84.4 ± 8.9, split slice-GRAPPA: 85.0 ± 8.9, ROCK-SPIRiT: 88.2 ± 6.6 [in percentage]). Regularized ROCK-SPIRiT significantly outperforms all methods, as characterized by these quantitative metrics (PSNR: 37.6 ± 3.8, SSIM: 94.2 ± 4.1 [in percentage]). The prospectively five-fold SMS and two-fold in-plane accelerated data show that ROCK-SPIRiT and regularized ROCK-SPIRiT have visually improved image quality compared with existing methods. CONCLUSION: The proposed ROCK-SPIRiT technique reduces noise and interslice leakage in accelerated SMS cardiac cine MRI, improving on existing methods both quantitatively and qualitatively.

Improved Simultaneous Multi-slice imaging with Composition of k-space Interpolations (SMS-COOKIE) for myocardial T1 mapping.

The aim of this study is to develop and evaluate a regularized Simultaneous Multi-Slice (SMS) reconstruction method for improved Cardiac Magnetic Resonance Imaging (CMR). The proposed reconstruction method, SMS with COmpOsition of k-space IntErpolations (SMS-COOKIE) combines the advantages of Iterative Self-consistent Parallel Imaging Reconstruction (SPIRiT) and split slice-Generalized Autocalibrating Partially Parallel Acquisitions (GRAPPA), while allowing regularization for further noise reduction. The proposed SMS-COOKIE was implemented with and without regularization, and validated using a Saturation Pulse-Prepared Heart rate Independent inversion REcovery (SAPPHIRE) myocardial T1 mapping sequence. The performance of the proposed reconstruction method was compared to ReadOut (RO)-SENSE-GRAPPA and split slice-GRAPPA, on both retrospectively and prospectively three-fold SMS-accelerated data with an additional two-fold in-plane acceleration. All SMS reconstruction methods yielded similar T1 values compared to single band imaging. SMS-COOKIE showed lower spatial variability in myocardial T1 with significant improvement over RO-SENSE-GRAPPA and split slice-GRAPPA (P < 10-4). The proposed method with additional locally low rank (LLR) regularization reduced the spatial variability, again with significant improvement over RO-SENSE-GRAPPA and split slice-GRAPPA (P < 10-4). In conclusion, improved reconstruction quality was achieved with the proposed SMS-COOKIE, which also provided lower spatial variability with significant improvement over split slice-GRAPPA.

High-fidelity Database-free Deep Learning Reconstruction for Real-time Cine Cardiac MRI.

Real-time cine cardiac MRI provides an ECG-free free-breathing alternative to clinical gold-standard ECG-gated breath-hold segmented cine MRI for evaluation of heart function. Real-time cine MRI data acquisition during free breathing snapshot imaging enables imaging of patient cohorts that cannot be imaged with segmented or breath-hold acquisitions, but requires rapid imaging to achieve sufficient spatial-temporal resolutions. However, at high acceleration rates, conventional reconstruction techniques suffer from residual aliasing and temporal blurring, including advanced methods such as compressed sensing with radial trajectories. Recently, deep learning (DL) reconstruction has emerged as a powerful tool in MRI. However, its utility for free-breathing real-time cine MRI has been limited, as database-learning of spatio-temporal correlations with varying breathing and cardiac motion patterns across subjects has been challenging. Zero-shot self-supervised physics-guided deep learning (PG-DL) reconstruction has been proposed to overcome such challenges of database training by enabling subject-specific training. In this work, we adapt zero-shot PG-DL for real-time cine MRI with a spatio-temporal regularization. We compare our method to TGRAPPA, locally low-rank (LLR) regularized reconstruction and database-trained PG-DL reconstruction, both for retrospectively and prospectively accelerated datasets. Results on highly accelerated real-time Cartesian cine MRI show that the proposed method outperforms other reconstruction methods, both visibly in terms of noise and aliasing, and quantitatively.

High-fidelity Database-free Deep Learning Reconstruction for Real-time Cine Cardiac MRI.

Real-time cine cardiac MRI provides an ECG-free free-breathing alternative to clinical gold-standard ECG-gated breath-hold segmented cine MRI for evaluation of heart function. Real-time cine MRI data acquisition during free breathing snapshot imaging enables imaging of patient cohorts that cannot be imaged with segmented or breath-hold acquisitions, but requires rapid imaging to achieve sufficient spatial-temporal resolutions. However, at high acceleration rates, conventional reconstruction techniques suffer from residual aliasing and temporal blurring, including advanced methods such as compressed sensing with radial trajectories. Recently, deep learning (DL) reconstruction has emerged as a powerful tool in MRI. However, its utility for free-breathing real-time cine MRI has been limited, as database-learning of spatio-temporal correlations with varying breathing and cardiac motion patterns across subjects has been challenging. Zero-shot self-supervised physics-guided deep learning (PG-DL) reconstruction has been proposed to overcome such challenges of database training by enabling subject-specific training. In this work, we adapt zero-shot PG-DL for real-time cine MRI with a spatio-temporal regularization. We compare our method to TGRAPPA, locally low-rank (LLR) regularized reconstruction and database-trained PG-DL reconstruction, both for retrospectively and prospectively accelerated datasets. Results on highly accelerated real-time Cartesian cine MRI show that the proposed method outperforms other reconstruction methods, both visibly in terms of noise and aliasing, and quantitatively.

Signal-Intensity Informed Multi-Coil MRI Encoding Operator for Improved Physics-Guided Deep Learning Reconstruction of Dynamic Contrast-Enhanced MRI.

Dynamic contrast enhanced (DCE) MRI acquires a series of images following the administration of a contrast agent, and plays an important clinical role in diagnosing various diseases. DCE MRI typically necessitates rapid imaging to provide sufficient spatio-temporal resolution and coverage. Conventional MRI acceleration techniques exhibit limited image quality at such high acceleration rates. Recently, deep learning (DL) methods have gained interest for improving highly-accelerated MRI. However, DCE MRI series show substantial variations in SNR and contrast across images. This hinders the quality and generalizability of DL methods, when applied across time frames. In this study, we propose signal intensity informed multi-coil MRI encoding operator for improved DL reconstruction of DCE MRI. The output of the corresponding inverse problem for this forward operator leads to more uniform contrast across time frames, since the proposed operator captures signal intensity variations across time frames while not altering the coil sensitivities. Our results in perfusion cardiac MRI show that high-quality images are reconstructed at very high acceleration rates, with substantial improvement over existing methods.

Myocardial Approximate Spin-lock Dispersion Mapping using a Simultaneous T2 and TRAFF2 Mapping at 3T MRI.

Ischemic heart disease (IHD) is one of the leading causes of death worldwide. Myocardial infarction (MI) represents a third of all IHD cases, and cardiac magnetic resonance imaging (MRI) is often used to assess its damage to myocardial viability. Late gadolinium enhancement (LGE) is the current gold standard, but the use of gadolinium-based agents limits the clinical applicability in some patients. Spin-lock (SL) dispersion has recently been proposed as a promising non-contrast biomarker for the assessment of MI. However, at 3T, the required range of SL preparations acquired at different amplitudes suffers from specific absorption rate (SAR) limitations and off-resonance artifacts. Relaxation Along a Fictitious Field (RAFF) is an alternative to SL preparations with lower SAR requirements, while still sampling relaxation in the rotating frame. In this study, a single breath-hold simultaneous TRAFF2 and T2 mapping sequence is proposed for SL dispersion mapping at 3T. Excellent reproducibility (coefficient of variations lower than 10%) was achieved in phantom experiments, indicating good intrascan repeatability. The average myocardial TRAFF2, T2, and SL dispersion obtained with the proposed sequence (68.0±10.7 ms, 44.0±4.0 ms, and 0.4±0.2 ×10-4 s2, respectively) were comparable to the reference methods (62.7±11.7 ms, 41.2±2.4 ms, and 0.3±0.2x 10-4s2, respectively). High visual map quality, free of B0 and B1+ related artifacts, for T2, TRAFF2, and SL dispersion maps were obtained in phantoms and in vivo, suggesting promise in clinical use at 3T. Clinical relevance - and imaging promises non-contrast assessment of scar and focal fibrosis in a single breath-hold using approximate spin-lock dispersion mapping.

20-fold Accelerated 7T fMRI Using Referenceless Self-Supervised Deep Learning Reconstruction.

High spatial and temporal resolution across the whole brain is essential to accurately resolve neural activities in fMRI. Therefore, accelerated imaging techniques target improved coverage with high spatio-temporal resolution. Simultaneous multi-slice (SMS) imaging combined with in-plane acceleration are used in large studies that involve ultrahigh field fMRI, such as the Human Connectome Project. However, for even higher acceleration rates, these methods cannot be reliably utilized due to aliasing and noise artifacts. Deep learning (DL) reconstruction techniques have recently gained substantial interest for improving highly-accelerated MRI. Supervised learning of DL reconstructions generally requires fully-sampled training datasets, which is not available for high-resolution fMRI studies. To tackle this challenge, self-supervised learning has been proposed for training of DL reconstruction with only undersampled datasets, showing similar performance to supervised learning. In this study, we utilize a self-supervised physics-guided DL reconstruction on a 5-fold SMS and 4-fold in-plane accelerated 7T fMRI data. Our results show that our self-supervised DL reconstruction produce high-quality images at this 20-fold acceleration, substantially improving on existing methods, while showing similar functional precision and temporal effects in the subsequent analysis compared to a standard 10-fold accelerated acquisition.

Anatomical measurements correlate with individual magnetostimulation thresholds for kHz-range homogeneous magnetic fields.

PURPOSE: Magnetostimulation, also known as peripheral nerve stimulation (PNS), is the dominant safety constraint in magnetic resonance imaging (MRI) for the gradient magnetic fields that operate around 0.1-1 kHz, and for the homogeneous drive field in magnetic particle imaging (MPI) that operates around 10-150 kHz. Previous studies did not report correlations between anatomical measures and magnetostimulation thresholds for the gradient magnetic fields in MRI. In contrast, a strong linear correlation was shown between the thresholds and the inverse of body part size in MPI. Yet, the effects of other anatomical measures on the thresholds for the drive field remain unexplored. Here, we investigate the effects of fat percentage on magnetostimulation thresholds for kHz-range homogeneous magnetic fields such as the drive field in MPI, with the ultimate goal of predicting subject-specific thresholds based on simple anatomical measures. METHODS: Human subject experiments were performed on the upper arms of 10 healthy male subjects (age: 26 ± 2 yr) to determine magnetostimulation thresholds. Experiments were repeated three times for each subject, with brief resting periods between repetitions. Using a solenoidal magnetostimulation coil, a homogeneous magnetic field at 25 kHz with 100 ms pulse duration was applied at 4-s intervals, while the subject reported stimulation via a mouse click. To determine the thresholds, individual subject responses were fitted to a cumulative distribution function modeled by a sigmoid curve. Next, anatomical images of the upper arms of the subjects were acquired on a 3 T MRI scanner. A two-point Dixon method was used to obtain separate images of water and fat tissues, from which several anatomical measures were derived: the effective outer radius of the upper arm, the effective inner radius (i.e., the muscle radius), and fat percentage. Pearson's correlation coefficient was used to assess the relationship between the threshold and anatomical measures. This statistical analysis was repeated after factoring out the expected effects of body part size. An updated model for threshold prediction is provided, where in addition to scaling in proportion with the inverse of the outer radius, the threshold has an affine dependence on fat percentage. RESULTS: A strong linear correlation (r = 0.783, P < 0.008) was found between magnetostimulation threshold and fat percentage, and the correlation became stronger after factoring out the effects of outer radius (r = 0.839, P < 0.003). While considering body part size alone did not explain any significant variance in measured thresholds (P > 0.398), the updated model that also incorporates fat percentage yielded substantially improved threshold predictions with R 2  = 0.654 (P < 0.001). CONCLUSIONS: This work shows for the first time that fat percentage strongly correlates with magnetostimulation thresholds for kHz-range homogenous magnetic fields such as the drive field in MPI, and that the correlations get even stronger after factoring out the effects of body part size. These results have important practical implications for predicting subject-specific thresholds, which in turn can increase the performance of the drive field and improve image quality while remaining within the safety limits.

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.

Functional LGE Imaging: Cardiac Phase-Resolved Assessment of Focal Fibrosis.

Cardiac Magnetic Resonance Imaging (CMR) is a central tool for diagnosis of various ischemic and non-ischemic cardiomyopathies. CMR protocols commonly comprise assessment of functional properties using cardiac phase-resolved CINE MRI and characterization of myocardial viability using late gadolinium enhancement (LGE) imaging. Conventional LGE imaging requires inversion recovery preparation with a specific inversion time to null the healthy myocardium, which restricts the acquisition to a single cardiac phase. In turn, this necessitates separate scans for cardiac function and viability. In this work, we develop a new method for functional LGE imaging in a single breath-hold using a three-step approach: 1) ECG-triggered multi-contrast data is acquired for each cardiac phase, 2) semi-quantitative relaxation maps are generated, 3) LGE imaging contrast is synthesized based on the semi-quantitative maps. The proposed functional LGE method is evaluated in four healthy subject and 20 patients at 1.5T and 3T. Thorough suppression of the healthy myocardium, as well as 40-80ms temporal resolution are achieved, with no visually apparent temporal blurring at tissue interfaces. Functional LGE in patients with focal scar demonstrates robust hyperenhancement in the scar area throughout all cardiac phases, allowing for visual assessment of scar motility. The proposed technique bears the potential to simplify and speedup common cardiac imaging protocols, while enabling improved data fusion of functional and viability information for improved evaluation of CMR.

Calibration-Free Relaxation-Based Multi-Color Magnetic Particle Imaging.

Magnetic particle imaging (MPI) is a novel imaging modality with important potential applications, such as angiography, stem cell tracking, and cancer imaging. Recently, there have been efforts to increase the functionality of MPI via multi-color imaging methods that can distinguish the responses of different nanoparticles, or nanoparticles in different environmental conditions. The proposed techniques typically rely on extensive calibrations that capture the differences in the harmonic responses of the nanoparticles. In this paper, we propose a method to directly estimate the relaxation time constant of the nanoparticles from the MPI signal, which is then used to generate a multi-color relaxation map. The technique is based on the underlying mirror symmetry of the adiabatic MPI signal when the same region is scanned back and forth. We validate the proposed method via simulations, and via experiments on our in-house magnetic particle spectrometer setup at 10.8 kHz and our in-house MPI scanner at 9.7 kHz. Our results show that nanoparticles can be successfully distinguished with the proposed technique, without any calibration or prior knowledge about the nanoparticles.