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.

Evaluating the effect of denoising submillimeter auditory fMRI data with NORDIC.

Functional magnetic resonance imaging (fMRI) has emerged as an essential tool for exploring human brain function. Submillimeter fMRI, in particular, has emerged as a tool to study mesoscopic computations. The inherently low signal-to-noise ratio (SNR) at submillimeter resolutions warrants the use of denoising approaches tailored at reducing thermal noise - the dominant contributing noise component in high resolution fMRI. NORDIC PCA is one of such approaches, and has been benchmarked against other approaches in several applications. Here, we investigate the effects that two versions of NORDIC denoising have on auditory submillimeter data. As investigating auditory functional responses poses unique challenges, we anticipated that the benefit of this technique would be especially pronounced. Our results show that NORDIC denoising improves the detection sensitivity and the reliability of estimates in submillimeter auditory fMRI data. These effects can be explained by the reduction of the noise-induced signal variability. However, we also observed a reduction in the average response amplitude (percent signal), which may suggest that a small amount of signal was also removed. We conclude that, while evaluating the effects of the signal reduction induced by NORDIC may be necessary for each application, using NORDIC in high resolution auditory fMRI studies may be advantageous because of the large reduction in variability of the estimated responses.

Intravoxel incoherent motion (IVIM) detects femoral head ischemia in a piglet model of Legg-Calvé-Perthes disease.

There is a clinical need for alternatives to gadolinium contrast-enhanced magnetic resonance imaging (MRI) to facilitate early detection and assessment of femoral head ischemia in pediatric patients with Legg-Calvé-Perthes disease (LCPD), a juvenile form of idiopathic osteonecrosis of the femoral head. The purpose of this study was to determine if intravoxel incoherent motion (IVIM), a noncontrast-enhanced MRI method to simultaneously measure tissue perfusion and diffusion, can detect femoral head ischemia using a piglet model of LCPD. Twelve 6-week-old piglets underwent unilateral hip surgery to induce complete femoral head ischemia. The unoperated, contralateral femoral head served as a perfused control. The bilateral hips of the piglets were imaged in vivo at 3T MRI using IVIM and contrast-enhanced MRI 1 week after surgery. Median apparent diffusion coefficient (ADC) and IVIM parameters (diffusion coefficient: Ds; perfusion coefficient: Df; perfusion fraction: f; and perfusion flux: f*Df) were compared between regions of interest comprising the epiphyseal bone marrow of the ischemic and control femoral heads. Contrast-enhanced MRI confirmed complete femoral head ischemia in 11/12 piglets. IVIM perfusion fraction (f) and flux (f*Df) were significantly decreased in the ischemic versus control femoral heads: on average, f decreased 47 ± 27% (Δf = -0.055 ± 0.034; p = 0.0003) and f*Df decreased 50 ± 27% (Δf*Df = -0.59 ± 0.49 × 10-3 mm2/s; p = 0.0026). In contrast, IVIM diffusion coefficient (Ds) and ADC were significantly increased in the ischemic versus control femoral heads: on average, Ds increased 78 ± 21% (ΔDs = 0.60 ± 0.14 × 10-3 mm2/s; p < 0.0001) and ADC increased 60 ± 36% (ΔADC = 0.50 ± 0.23 × 10-3 mm2/s; p < 0.0001). In conclusion, IVIM is sensitive in detecting bone marrow ischemia in a piglet model of LCPD.

Locally low-rank denoising in transform domains.

Purpose: To develop an extension to locally low rank (LLR) denoising techniques based on transform domain processing that reduces the number of images required in the MR image series for high-quality denoising. Theory and Methods: LLR methods with random matrix theory-based thresholds are successfully used in the denoising of MR image series in a number of applications. The performance of these methods depend on how well the LLR assumption is satisfied, which deteriorates with few numbers of images, as is commonly encountered in quantitative MRI applications. We propose a transform-domain approach for denoising of MR image series to represent the underlying signal with higher fidelity when using a locally low rank approximation. The efficacy of the method is demonstrated for fully-sampled k-space, undersampled k-space, DICOM images, and complex-valued SENSE-1 images in quantitative MRI applications with as few as 4 images. Results: For both MSK and brain applications, the transform domain denoising preserves local subtle variability, whereas the quantitative maps based on image domain LLR methods tend to be locally more homogeneous. Conclusion: A transform domain extension to LLR denoising produces high quality images and is compatible with both raw k-space data and vendor reconstructed data. This allows for improved imaging and more accurate quantitative analyses and parameters obtained therefrom.

DENOISING DIFFUSION MRI: CONSIDERATIONS AND IMPLICATIONS FOR ANALYSIS.

Development of diffusion MRI (dMRI) denoising approaches has experienced considerable growth over the last years. As noise can inherently reduce accuracy and precision in measurements, its effects have been well characterised both in terms of uncertainty increase in dMRI-derived features and in terms of biases caused by the noise floor, the smallest measurable signal given the noise level. However, gaps in our knowledge still exist in objectively characterising dMRI denoising approaches in terms of both of these effects and assessing their efficacy. In this work, we reconsider what a denoising method should and should not do and we accordingly define criteria to characterise the performance. We propose a comprehensive set of evaluations, including i) benefits in improving signal quality and reducing noise variance, ii) gains in reducing biases and the noise floor and improving, iii) preservation of spatial resolution, iv) agreement of denoised data against a gold standard, v) gains in downstream parameter estimation (precision and accuracy), vi) efficacy in enabling noise-prone applications, such as ultra-high-resolution imaging. We further provide newly acquired complex datasets (magnitude and phase) with multiple repeats that sample different SNR regimes to highlight performance differences under different scenarios. Without loss of generality, we subsequently apply a number of exemplar patch-based denoising algorithms to these datasets, including Non-Local Means, Marchenko-Pastur PCA (MPPCA) in the magnitude and complex domain and NORDIC, and compare them with respect to the above criteria and against a gold standard complex average of multiple repeats. We demonstrate that all tested denoising approaches reduce noise-related variance, but not always biases from the elevated noise floor. They all induce a spatial resolution penalty, but its extent can vary depending on the method and the implementation. Some denoising approaches agree with the gold standard more than others and we demonstrate challenges in even defining such a standard. Overall, we show that dMRI denoising performed in the complex domain is advantageous to magnitude domain denoising with respect to all the above criteria.

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.

Evaluating increases in sensitivity from NORDIC for diverse fMRI acquisition strategies.

As the neuroimaging field moves towards detecting smaller effects at higher spatial resolutions, and faster sampling rates, there is increased attention given to the deleterious contribution of unstructured, thermal noise. Here, we critically evaluate the performance of a recently developed reconstruction method, termed NORDIC, for suppressing thermal noise using datasets acquired with various field strengths, voxel sizes, sampling rates, and task designs. Following minimal preprocessing, statistical activation (t-values) of NORDIC processed data was compared to the results obtained with alternative denoising methods. Additionally, we examined the consistency of the estimates of task responses at the single-voxel, single run level, using a finite impulse response (FIR) model. To examine the potential impact on effective image resolution, the overall smoothness of the data processed with different methods was estimated. Finally, to determine if NORDIC alters or removes temporal information important for modeling responses, we employed an exhaustive leave-p-out cross validation approach, using FIR task responses to predict held out timeseries, quantified using R2. After NORDIC, the t-values are increased, an improvement comparable to what could be achieved by 1.5 voxels smoothing, and task events are clearly visible and have less cross-run error. These advantages are achieved with smoothness estimates increasing by less than 4%, while 1.5 voxel smoothing is associated with increases of over 140%. Cross-validated R2s based on the FIR models show that NORDIC is not measurably distorting the temporal structure of the data under this approach and is the best predictor of non-denoised time courses. The results demonstrate that analyzing 1 run of data after NORDIC produces results equivalent to using 2 to 3 original runs and that NORDIC performs equally well across a diverse array of functional imaging protocols. Significance Statement: For functional neuroimaging, the increasing availability of higher field strengths and ever higher spatiotemporal resolutions has led to concomitant increase in concerns about the deleterious effects of thermal noise. Historically this noise source was suppressed using methods that reduce spatial precision such as image blurring or averaging over a large number of trials or sessions, which necessitates large data collection efforts. Here, we critically evaluate the performance of a recently developed reconstruction method, termed NORDIC, which suppresses thermal noise. Across datasets varying in field strength, voxel sizes, sampling rates, and task designs, NORDIC produces substantial gains in data quality. Both conventional t-statistics derived from general linear models and coefficients of determination for predicting unseen data are improved. These gains match or even exceed those associated with 1 voxel Full Width Half Max image smoothing, however, even such small amounts of smoothing are associated with a 52% reduction in estimates of spatial precision, whereas the measurable difference in spatial precision is less than 4% following NORDIC.

Motion robust magnetic resonance imaging via efficient Fourier aggregation.

We present a method for suppressing motion artifacts in anatomical magnetic resonance acquisitions. Our proposed technique, termed MOTOR-MRI, can recover and salvage images which are otherwise heavily corrupted by motion induced artifacts and blur which renders them unusable. Contrary to other techniques, MOTOR-MRI operates on the reconstructed images and not on k-space data. It relies on breaking the standard acquisition protocol into several shorter ones (while maintaining the same total acquisition time) and subsequent efficient aggregation in Fourier space of locally sharp and consistent information among them, producing a sharp and motion mitigated image. We demonstrate the efficacy of the technique on T2-weighted turbo spin echo magnetic resonance brain scans with severe motion corruption from both 3 T and 7 T scanners and show significant qualitative and quantitative improvement in image quality. MOTOR-MRI can operate independently, or in conjunction with additional motion correction methods.

Evaluation of lesion and overlying articular cartilage in patients with juvenile osteochondritis dissecans of the knee using quantitative diffusion MRI.

Current clinical MRI of patients with juvenile osteochondritis dissecans (JOCD) is limited by the low reproducibility of lesion instability evaluation and inability to predict which lesions will heal after nonoperative treatment and which will later require surgery. The aim of this study is to verify the ability of apparent diffusion coefficient (ADC) to detect differences in lesion microstructure between different JOCD stages, treatment groups, and healthy, unaffected contralateral knees. Pediatric patients with JOCD received quantitative diffusion MRI between January 2016 and September 2020 in this prospective research study. A disease stage (I-IV) and stability of each JOCD lesion was evaluated. ADCs were calculated in progeny lesion, interface, parent bone, cartilage overlying lesion, control bone, and control cartilage regions. ADC differences were evaluated using linear mixed models with Bonferroni correction. Evaluated were 30 patients (mean age, 13 years; 21 males), with 40 JOCD-affected and 12 healthy knees. Nine patients received surgical treatment after MRI. Negative Spearman rank correlations were found between ADCs and JOCD stage in the progeny lesion (ρ = -0.572; p < 0.001), interface (ρ = -0.324; p = 0.041), and parent bone (ρ = -0.610; p < 0.001), demonstrating the sensitivity of ADC to microstructural differences in lesions at different JOCD stages. We observed a significant increase in the interface ADCs (p = 0.007) between operative (mean [95% CI] = 1.79 [1.56-2.01] × 10-3 mm2 /s) and nonoperative group (1.27 [0.98-1.57] × 10-3 mm2 /s). Quantitative diffusion MRI detects microstructural differences in lesions at different stages of JOCD progression towards healing and reveals differences between patients assigned for operative versus nonoperative treatment.

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.

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.

NOise Reduction with DIstribution Corrected (NORDIC) PCA improves signal-to-noise in rodent resting-state and optogenetic functional MRI.

NOise Reduction with DIstribution Corrected (NORDIC) principal component analysis (PCA) has been shown to selectively suppress thermal noise and improve temporal signal-to-noise ratio (tSNR) in human functional magnetic resonance imaging (fMRI). However, the feasibility to improve rodent fMRI using NORDIC PCA has not been explored. In this study, we developed a rodent fMRI preprocessing pipeline by incorporating NORDIC and evaluated its performance in a range of rodent fMRI applications from resting-state fMRI to task-evoked fMRI using optogenetics. In resting-state fMRI, we demonstrated a significant increase in tSNR by more than 3 times after NORDIC correction with reduced variance and improved task-free relative cerebrovascular reactivity (rCVR) across cortical depth. In optogenetic fMRI, apart from tSNR increase, more activated voxels and a significant decrease in the variance of activated brain signals were observed after NORDIC correction without apparent change in brain morphology. Taken together, our results signified the values of NORDIC correction for better detection of brain activities in rodent fMRI. Clinical Relevance: NORDIC PCA increases temporal signalto- noise ratio in rodent resting-state and task-evoked functional MRI, which can play an important role in improving the image quality for translational medicine and preclinical research, and for guiding future clinical neuroimaging.

Revisiting [Formula: see text]-wavelet compressed-sensing MRI in the era of deep learning.

Following their success in numerous imaging and computer vision applications, deep-learning (DL) techniques have emerged as one of the most prominent strategies for accelerated MRI reconstruction. These methods have been shown to outperform conventional regularized methods based on compressed sensing (CS). However, in most comparisons, CS is implemented with two or three hand-tuned parameters, while DL methods enjoy a plethora of advanced data science tools. In this work, we revisit [Formula: see text]-wavelet CS reconstruction using these modern tools. Using ideas such as algorithm unrolling and advanced optimization methods over large databases that DL algorithms utilize, along with conventional insights from wavelet representations and CS theory, we show that [Formula: see text]-wavelet CS can be fine-tuned to a level close to DL reconstruction for accelerated MRI. The optimized [Formula: see text]-wavelet CS method uses only 128 parameters compared to >500,000 for DL, employs a convex reconstruction at inference time, and performs within <1% of a DL approach that has been used in multiple studies in terms of quantitative quality metrics.

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 TSE imaging at ultrahigh field using nonlocalized efficiency RF shimming and acquisition modes optimized for refocused echoes (AMORE).

PURPOSE: To develop and evaluate a novel RF shimming optimization strategy tailored to improve the transmit efficiency in turbo spin echo imaging when performing time-interleaved acquisition of modes (TIAMO) at ultrahigh fields. THEORY AND METHODS: A nonlocalized efficiency shimming cost function is proposed and extended to perform TIAMO using acquisition modes optimized for refocused echoes (AMORE). The nonlocalized efficiency shimming was demonstrated in brain and knee imaging at 7 Tesla. Phantom and in vivo torso imaging studies were performed to compare the performance between AMORE and previously proposed TIAMO mode optimizations with and without localized constraints in turbo spin echo and gradient echo acquisitions. RESULTS: The proposed nonlocalized efficiency RF shimming produced a circularly polarized-like field with fewer signal dropouts in the brain and knee. For larger targets, AMORE was used and required a significantly lower transmitter voltage to produce a similar contrast to existing TIAMO mode design approaches for turbo spin echo as well as gradient echo acquisitions. In vivo, AMORE effectively reduced signal dropout in the interior torso while providing more uniform contrast with reduced transmit power. A local constraint further improved performance for a target region while maintaining performance in the larger FOV. CONCLUSION: AMORE based on the presented nonlocalized efficiency shimming cost function demonstrated improved contrast and SNR uniformity as well as increased transmit efficiency for both gradient echo and turbo spin echo acquisitions.

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.

Compressed Sensing MRI with ℓ1-Wavelet Reconstruction Revisited Using Modern Data Science Tools.

Deep learning (DL) has emerged as a powerful tool for improving the reconstruction quality of accelerated MRI. These methods usually show enhanced performance compared to conventional methods, such as compressed sensing (CS) and parallel imaging. However, in most scenarios, CS is implemented with two or three empirically-tuned hyperparameters, while a plethora of advanced data science tools are used in DL. In this work, we revisit ℓ1 -wavelet CS for accelerated MRI using modern data science tools. By using tools like algorithm unrolling and end-to-end training with stochastic gradient descent over large databases that DL algorithms utilize, and combining these with conventional concepts like wavelet sub-band processing and reweighted ℓ1 minimization, we show that ℓ1-wavelet CS can be fine-tuned to a level comparable to DL methods. While DL uses hundreds of thousands of parameters, the proposed optimized ℓ1-wavelet CS with sub-band training and reweighting uses only 128 parameters, and employs a fully-explainable convex reconstruction model.

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.

Lowering the thermal noise barrier in functional brain mapping with magnetic resonance imaging.

Functional magnetic resonance imaging (fMRI) has become an indispensable tool for investigating the human brain. However, the inherently poor signal-to-noise-ratio (SNR) of the fMRI measurement represents a major barrier to expanding its spatiotemporal scale as well as its utility and ultimate impact. Here we introduce a denoising technique that selectively suppresses the thermal noise contribution to the fMRI experiment. Using 7-Tesla, high-resolution human brain data, we demonstrate improvements in key metrics of functional mapping (temporal-SNR, the detection and reproducibility of stimulus-induced signal changes, and accuracy of functional maps) while leaving the amplitude of the stimulus-induced signal changes, spatial precision, and functional point-spread-function unaltered. We demonstrate that the method enables the acquisition of ultrahigh resolution (0.5 mm isotropic) functional maps but is also equally beneficial for a large variety of fMRI applications, including supra-millimeter resolution 3- and 7-Tesla data obtained over different cortical regions with different stimulation/task paradigms and acquisition strategies.

A self-decoupled 32-channel receive array for human-brain MRI at 10.5 T.

PURPOSE: Receive array layout, noise mitigation, and B0 field strength are crucial contributors to SNR and parallel-imaging performance. Here, we investigate SNR and parallel-imaging gains at 10.5 T compared with 7 T using 32-channel receive arrays at both fields. METHODS: A self-decoupled 32-channel receive array for human brain imaging at 10.5 T (10.5T-32Rx), consisting of 31 loops and one cloverleaf element, was co-designed and built in tandem with a 16-channel dual-row loop transmitter. Novel receive array design and self-decoupling techniques were implemented. Parallel imaging performance, in terms of SNR and noise amplification (g-factor), of the 10.5T-32Rx was compared with the performance of an industry-standard 32-channel receiver at 7 T (7T-32Rx) through experimental phantom measurements. RESULTS: Compared with the 7T-32Rx, the 10.5T-32Rx provided 1.46 times the central SNR and 2.08 times the peripheral SNR. Minimum inverse g-factor value of the 10.5T-32Rx (min[1/g] = 0.56) was 51% higher than that of the 7T-32Rx (min[1/g] = 0.37) with R = 4 × 4 2D acceleration, resulting in significantly enhanced parallel-imaging performance at 10.5 T compared with 7 T. The g-factor values of 10.5 T-32 Rx were on par with those of a 64-channel receiver at 7 T (eg, 1.8 vs 1.9, respectively, with R = 4 × 4 axial acceleration). CONCLUSION: Experimental measurements demonstrated effective self-decoupling of the receive array as well as substantial gains in SNR and parallel-imaging performance at 10.5 T compared with 7 T.