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

Knowledge-driven deep learning for fast MR imaging: Undersampled MR image reconstruction from supervised to un-supervised learning

Deep learning (DL) has emerged as a leading approach in accelerating MRI. It employs deep neural networks to extract knowledge from available datasets and then applies the trained networks to reconstruct accurate images from limited measurements. Unlike natural image restoration problems, MRI involves physics-based imaging processes, unique data properties, and diverse imaging tasks. This domain knowledge needs to be integrated with data-driven approaches. Our review will introduce the significant challenges faced by such knowledge-driven DL approaches in the context of fast MRI along with several notable solutions, which include learning neural networks and addressing different imaging application scenarios. The traits and trends of these techniques have also been given which have shifted from supervised learning to semi-supervised learning, and finally, to unsupervised learning methods. In addition, MR vendors' choices of DL reconstruction have been provided along with some discussions on open questions and future directions, which are critical for the reliable imaging systems.

Accelerated multi-b-value multi-shot diffusion-weighted imaging based on EPI with keyhole and a low-rank tensor constraint

PURPOSE

Multi-Shot (MS) Echo-Planar Imaging (EPI) may improve the in-plane resolution of multi-b-value DWI, yet it also considerably increases the scan time. Here we explored the combination of EPI with Keyhole (EPIK) and a calibrationless reconstruction algorithm for acceleration of multi-b-value MS-DWI.

METHODS

We firstly analyzed the impact of nonuniform phase accrual in EPIK on the reconstructed image. Based on insights gained from the analysis, we developed a calibrationless reconstruction algorithm based on a Space-Contrast-Coil Locally Low-Rank Tensor (SCC-LLRT) constraint for reconstruction of EPIK-acquired data. We compared the algorithm with a modified SPatial-Angular Locally Low-Rank (SPA-LLR) algorithm through simulations, phantoms, and in vivo study. We then compared EPIK with uniformly undersampled EPI for accelerating multi-b-value DWI in 6 healthy subjects.

RESULTS

Through theoretical derivations, we found that the reconstruction of EPIK with a SENSE-encoding-based algorithm, such as SPA-LLR, may cause additional aliasing artifacts due to the frequency-dependent distortion of the coil sensitivity. Results from simulations, phantoms, and in vivo study verified the theoretical finding by showing that the calibrationless SCC-LLRT algorithm reduced aliasing artifacts compared with SPA-LLR. Finally, EPIK with SCC-LLRT substantially reduced the ghosting artifacts compared with uniform undersampled multi-b-value DWI, decreasing the fitting errors in ADC (0.05 ± 0.01 vs 0.10 ± 0.01, P < 0.001) and IVIM mapping (0.026 ± 0.004 vs 0.06 ± 0.006, P < 0.001).

CONCLUSION

The SCC-LLRT algorithm reduced the aliasing artifacts of EPIK by using a calibrationless modeling of the multi-coil data. The dense sampling of k-space center offers EPIK a potential to improve image quality for acceleration of multi-b-value MS-DWI.

Integrating data distribution prior via Langevin dynamics for end-to-end MR reconstruction

PURPOSE

To develop a novel deep learning-based method inheriting the advantages of data distribution prior and end-to-end training for accelerating MRI.

METHODS

Langevin dynamics is used to formulate image reconstruction with data distribution before facilitate image reconstruction. The data distribution prior is learned implicitly through the end-to-end adversarial training to mitigate the hyper-parameter selection and shorten the testing time compared to traditional probabilistic reconstruction. By seamlessly integrating the deep equilibrium model, the iteration of Langevin dynamics culminates in convergence to a fix-point, ensuring the stability of the learned distribution.

RESULTS

The feasibility of the proposed method is evaluated on the brain and knee datasets. Retrospective results with uniform and random masks show that the proposed method demonstrates superior performance both quantitatively and qualitatively than the state-of-the-art.

CONCLUSION

The proposed method incorporating Langevin dynamics with end-to-end adversarial training facilitates efficient and robust reconstruction for MRI. Empirical evaluations conducted on brain and knee datasets compellingly demonstrate the superior performance of the proposed method in terms of artifact removing and detail preserving.

Improved temporal resolution and acceleration on 4D-MR angiography based on superselective pseudo-continuous arterial spin labeling combined with CENTRA-keyhole and view-sharing (4D-S-PACK) using an interpolation algorithm on the temporal axis and compressed sensing-sensitivity encoding (CS-SENSE)

PURPOSE

Two major drawbacks of 4D-MR angiography based on superselective pseudo-continuous arterial spin labeling combined with CENTRA-keyhole and view-sharing (4D-S-PACK) are the low temporal resolution and long scanning time. We investigated the feasibility of increasing the temporal resolution and accelerating the scanning time on 4D-S-PACK by using CS-SENSE and PhyZiodynamics, a novel image-processing program that interpolates images between phases to generate new phases and reduces image noise.

METHODS

Seven healthy volunteers were scanned with a 3.0 T MR scanner to visualize the internal carotid artery (ICA) system. PhyZiodynamics is a novel image-processing that interpolates images between phases to generate new phases and reduces image noise, and by increasing temporal resolution using PhyZiodynamics, inflow dynamic data (reference) were acquired by changing the labeling durations (100-2000 msec, 31 phases) in 4D-S-PACK. From this set of data, we selected seven time intervals to calculate interpolated time points with up to 61 intervals using ×10 for the generation of interpolated phases with PhyZiodynamics. In the denoising process of PhyZiodynamics, we processed the none, low, medium, high noise reduction dataset images. The time intensity curve (TIC), the contrast-to-noise ratio (CNR) were evaluated. In accelerating with CS-SENSE for 4D-S-PACK, 4D-S-PACK were scanned different SENSE or CS-SENSE acceleration factors: SENSE3, CS3-6. Signal intensity (SI), CNR, were evaluated for accelerating the 4D-S-PACK. With regard to arterial vascular visualization, we evaluated the middle cerebral artery (MCA: M1-4 segments).

RESULTS

In increasing temporal resolution, the TIC showed a similar trend between the reference dataset and the interpolated dataset. As the noise reduction weight increased, the CNR of the interpolated dataset were increased compared to that of the reference dataset. In accelerating 4D-S-PACK, the SI values of the SENSE3 dataset and CS dataset with CS3-6 were no significant differences. The image noise increased with the increase of acceleration factor, and the CNR decreased with the increase of acceleration factor. Significant differences in CNR were observed between acceleration factor of SENSE3 and CS6 for the M1-4 (P < 0.05). Visualization of small arteries (M4) became less reliable in CS5 or CS6 images. Significant differences were found for the scores of M2, M3 and M4 segments between SENSE3 and CS6.

CONCLUSION

With PhyZiodynamics and CS-SENSE in 4D-S-PACK, we were able to shorten the scan time while improving the temporal resolution.

A noise robust image reconstruction using slice aware cycle interpolator network for parallel imaging in MRI

BACKGROUND

Reducing Magnetic resonance imaging (MRI) scan time has been an important issue for clinical applications. In order to reduce MRI scan time, imaging acceleration was made possible by undersampling k-space data. This is achieved by leveraging additional spatial information from multiple, independent receiver coils, thereby reducing the number of sampled k-space lines.

PURPOSE

The aim of this study is to develop a deep-learning method for parallel imaging with a reduced number of auto-calibration signals (ACS) lines in noisy environments.

METHODS

A cycle interpolator network is developed for robust reconstruction of parallel MRI with a small number of ACS lines in noisy environments. The network estimates missing (unsampled) lines of each coil data, and these estimated missing lines are then utilized to re-estimate the sampled k-space lines. In addition, a slice aware reconstruction technique is developed for noise-robust reconstruction while reducing the number of ACS lines. We conducted an evaluation study using retrospectively subsampled data obtained from three healthy volunteers at 3T MRI, involving three different slice thicknesses (1.5, 3.0, and 4.5 mm) and three different image contrasts (T1w, T2w, and FLAIR).

RESULTS

Despite the challenges posed by substantial noise in cases with a limited number of ACS lines and thinner slices, the slice aware cycle interpolator network reconstructs the enhanced parallel images. It outperforms RAKI, effectively eliminating aliasing artifacts. Moreover, the proposed network outperforms GRAPPA and demonstrates the ability to successfully reconstruct brain images even under severe noisy conditions.

CONCLUSIONS

The slice aware cycle interpolator network has the potential to improve reconstruction accuracy for a reduced number of ACS lines in noisy environments.

MRI reconstruction with enhanced self-similarity using graph convolutional network

BACKGROUND

Recent Convolutional Neural Networks (CNNs) perform low-error reconstruction in fast Magnetic Resonance Imaging (MRI). Most of them convolve the image with kernels and successfully explore the local information. Nonetheless, the non-local image information, which is embedded among image patches relatively far from each other, may be lost due to the limitation of the receptive field of the convolution kernel. We aim to incorporate a graph to represent non-local information and improve the reconstructed images by using the Graph Convolutional Enhanced Self-Similarity (GCESS) network.

METHODS

First, the image is reconstructed into the graph to extract the non-local self-similarity in the image. Second, GCESS uses spatial convolution and graph convolution to process the information in the image, so that local and non-local information can be effectively utilized. The network strengthens the non-local similarity between similar image patches while reconstructing images, making the reconstruction of structure more reliable.

RESULTS

Experimental results on in vivo knee and brain data demonstrate that the proposed method achieves better artifact suppression and detail preservation than state-of-the-art methods, both visually and quantitatively. Under 1D Cartesian sampling with 4 × acceleration (AF = 4), the PSNR of knee data reached 34.19 dB, 1.05 dB higher than that of the compared methods; the SSIM achieved 0.8994, 2% higher than the compared methods. Similar results were obtained for the reconstructed images under other sampling templates as demonstrated in our experiment.

CONCLUSIONS

The proposed method successfully constructs a hybrid graph convolution and spatial convolution network to reconstruct images. This method, through its training process, amplifies the non-local self-similarities, significantly benefiting the structural integrity of the reconstructed images. Experiments demonstrate that the proposed method outperforms the state-of-the-art reconstruction method in suppressing artifacts, as well as in preserving image details.

Non-Cartesian 3D-SPARKLING vs Cartesian 3D-EPI encoding schemes for functional Magnetic Resonance Imaging at 7 Tesla

The quest for higher spatial and/or temporal resolution in functional MRI (fMRI) while preserving a sufficient temporal signal-to-noise ratio (tSNR) has generated a tremendous amount of methodological contributions in the last decade ranging from Cartesian vs. non-Cartesian readouts, 2D vs. 3D acquisition strategies, parallel imaging and/or compressed sensing (CS) accelerations and simultaneous multi-slice acquisitions to cite a few. In this paper, we investigate the use of a finely tuned version of 3D-SPARKLING. This is a non-Cartesian CS-based acquisition technique for high spatial resolution whole-brain fMRI. We compare it to state-of-the-art Cartesian 3D-EPI during both a retinotopic mapping paradigm and resting-state acquisitions at 1mm3 (isotropic spatial resolution). This study involves six healthy volunteers and both acquisition sequences were run on each individual in a randomly-balanced order across subjects. The performances of both acquisition techniques are compared to each other in regards to tSNR, sensitivity to the BOLD effect and spatial specificity. Our findings reveal that 3D-SPARKLING has a higher tSNR than 3D-EPI, an improved sensitivity to detect the BOLD contrast in the gray matter, and an improved spatial specificity. Compared to 3D-EPI, 3D-SPARKLING yields, on average, 7% more activated voxels in the gray matter relative to the total number of activated voxels.

Encoding Enhanced Complex CNN for Accurate and Highly Accelerated MRI

Magnetic resonance imaging (MRI) using hyperpolarized noble gases provides a way to visualize the structure and function of human lung, but the long imaging time limits its broad research and clinical applications. Deep learning has demonstrated great potential for accelerating MRI by reconstructing images from undersampled data. However, most existing deep convolutional neural networks (CNN) directly apply square convolution to k-space data without considering the inherent properties of k-space sampling, limiting k-space learning efficiency and image reconstruction quality. In this work, we propose an encoding enhanced (EN2) complex CNN for highly undersampled pulmonary MRI reconstruction. EN2 complex CNN employs convolution along either the frequency or phase-encoding direction, resembling the mechanisms of k-space sampling, to maximize the utilization of the encoding correlation and integrity within a row or column of k-space. We also employ complex convolution to learn rich representations from the complex k-space data. In addition, we develop a feature-strengthened modularized unit to further boost the reconstruction performance. Experiments demonstrate that our approach can accurately reconstruct hyperpolarized 129Xe and 1H lung MRI from 6-fold undersampled k-space data and provide lung function measurements with minimal biases compared with fully sampled images. These results demonstrate the effectiveness of the proposed algorithmic components and indicate that the proposed approach could be used for accelerated pulmonary MRI in research and clinical lung disease patient care.

High-Frequency Space Diffusion Model for Accelerated MRI

Diffusion models with continuous stochastic differential equations (SDEs) have shown superior performances in image generation. It can serve as a deep generative prior to solving the inverse problem in magnetic resonance (MR) reconstruction. However, low-frequency regions of k -space data are typically fully sampled in fast MR imaging, while existing diffusion models are performed throughout the entire image or k -space, inevitably introducing uncertainty in the reconstruction of low-frequency regions. Additionally, existing diffusion models often demand substantial iterations to converge, resulting in time-consuming reconstructions. To address these challenges, we propose a novel SDE tailored specifically for MR reconstruction with the diffusion process in high-frequency space (referred to as HFS-SDE). This approach ensures determinism in the fully sampled low-frequency regions and accelerates the sampling procedure of reverse diffusion. Experiments conducted on the publicly available fastMRI dataset demonstrate that the proposed HFS-SDE method outperforms traditional parallel imaging methods, supervised deep learning, and existing diffusion models in terms of reconstruction accuracy and stability. The fast convergence properties are also confirmed through theoretical and experimental validation. Our code and weights are available at https://github.com/Aboriginer/HFS-SDE.

Variable augmentation network for invertible MR coil compression

To improve the efficiency of multi-coil data compression and recover the compressed image reversibly, increasing the possibility of applying the proposed method to medical scenarios. A deep learning algorithm is employed for MR coil compression in the presented work. The approach introduces a variable augmentation network for invertible coil compression (VAN-ICC). This network utilizes the inherent reversibility of normalizing flow-based models. The aim is to enhance the readability of the sentence and clearly convey the key components of the algorithm. By applying the variable augmentation technology to image/k-space variables from multi-coils, VAN-ICC trains the invertible network by finding an invertible and bijective function, which can map the original data to the compressed counterpart and vice versa. Experiments conducted on both fully-sampled and under-sampled data verified the effectiveness and flexibility of VAN-ICC. Quantitative and qualitative comparisons with traditional non-deep learning-based approaches demonstrated that VAN-ICC carries much higher compression effects. The proposed method trains the invertible network by finding an invertible and bijective function, which improves the defects of traditional coil compression method by utilizing inherent reversibility of normalizing flow-based models. In addition, the application of variable augmentation technology ensures the implementation of reversible networks. In short, VAN-ICC offered a competitive advantage over other traditional coil compression algorithms.

Single-shot frequency offset measurement with HASTE using the selective parity approach

Measurements of frequency offset are commonly required in MRI. The standard method measures the signal phase as a function of evolution time. Here we use a single shot turbo-spin-echo acquisition method to measure frequency offset at a single evolution time. After excitation the transverse magnetisation evolves during the evolution time, and is then repeatedly refocused. The phase is conjugated between alternate echoes. Using partial parallel acquisition techniques we obtain separate odd- and even- echo images. An iterative procedure ensures self-consistency between them. The difference in phase between the two images yields frequency offset maps. The technique was implemented at 3 Tesla and tested on a healthy human volunteer for a range of evolution times between 6 and 42 ms. A standard method using a similar readout train and multiple evolution times was used as a gold-standard measure. In a statistical comparison with the gold standard no evidence for bias or offset was found. There was no systematic variation in precision or accuracy as a function of evolution time. We conclude that the presented approach represents a viable method for the rapid generation of frequency offset maps with a high image quality and minimal distortion.

Feasibility of undersampled spiral trajectories in MREPT for fast conductivity imaging

PURPOSE

To investigate spiral-based imaging including trajectories with undersampling as a fast and robust alternative for phase-based magnetic resonance electrical properties tomography (MREPT) techniques.

METHODS

Spiral trajectories with various undersampling ratios were prescribed to acquire images from an experimental phantom and a healthy volunteer at 3T. The non-Cartesian acquisitions were reconstructed using SPIRiT, and conductivity maps were derived using phase-based cr-MREPT. The resulting maps were compared between different sampling trajectories. Additionally, a conductivity map was obtained using a Cartesian balanced SSFP acquisition from the volunteer to comparatively demonstrate the robustness of the proposed method.

RESULTS

The phantom and volunteer results illustrate the benefits of the spiral acquisitions. Specifically, undersampled spiral acquisitions display improved robustness against field inhomogeneity artifacts and lowered SD values with shortened readout times. Furthermore, average of conductivity values measured for the cerebrospinal fluid with the spiral acquisitions were 1.703 S/m, indicating a close agreement with the theoretical values of 1.794 S/m.

CONCLUSION

A spiral-based acquisition framework for conductivity imaging with and without undersampling is presented. Overall, spiral-based acquisitions improved robustness against field inhomogeneity artifacts, while achieving whole head coverage with multiple averages in less than a minute.

High efficiency free-breathing 3D thoracic aorta vessel wall imaging using self-gating image reconstruction

PURPOSE

To improve the scan efficiency of thoracic aorta vessel wall imaging using a self-gating (SG)-based motion correction scheme.

MATERIALS AND METHODS

A slab-selective variable-flip-angle 3D turbo spin-echo (SPACE) sequence was modified to acquire SG signals and imaging data. Cartesian sampling with a tiny golden-step spiral profile ordering was used to obtain the imaging data during the systolic period, and then the image data were subsequently corrected based on the SG signals and binned to different respiratory cycles. Finally, respiratory artifacts were estimated from image-based registration of 3D undersampled respiratory bins that were reconstructed with L1 iterative self-consistent parallel imaging reconstruction (SPIRiT). This method was evaluated in 11 healthy volunteers and compared against conventional diaphragmatic navigator-gated acquisition to assess the feasibility of the proposed framework.

RESULTS

Results showed that the proposed method achieved image quality comparable to that of conventional diaphragmatic navigator-gated acquisition with an average scan time of 4 min. The sharpness of the vessel wall and the definition of the liver boundary were in good agreement with the navigator-gated acquisition, which took approximately above 8.5 min depend on the respiratory rate. Further valuation of this technique in patients will be conducted to determine its clinical use.

IMJENSE: Scan-Specific Implicit Representation for Joint Coil Sensitivity and Image Estimation in Parallel MRI

Parallel imaging is a commonly used technique to accelerate magnetic resonance imaging (MRI) data acquisition. Mathematically, parallel MRI reconstruction can be formulated as an inverse problem relating the sparsely sampled k-space measurements to the desired MRI image. Despite the success of many existing reconstruction algorithms, it remains a challenge to reliably reconstruct a high-quality image from highly reduced k-space measurements. Recently, implicit neural representation has emerged as a powerful paradigm to exploit the internal information and the physics of partially acquired data to generate the desired object. In this study, we introduced IMJENSE, a scan-specific implicit neural representation-based method for improving parallel MRI reconstruction. Specifically, the underlying MRI image and coil sensitivities were modeled as continuous functions of spatial coordinates, parameterized by neural networks and polynomials, respectively. The weights in the networks and coefficients in the polynomials were simultaneously learned directly from sparsely acquired k-space measurements, without fully sampled ground truth data for training. Benefiting from the powerful continuous representation and joint estimation of the MRI image and coil sensitivities, IMJENSE outperforms conventional image or k-space domain reconstruction algorithms. With extremely limited calibration data, IMJENSE is more stable than supervised calibrationless and calibration-based deep-learning methods. Results show that IMJENSE robustly reconstructs the images acquired at 5× and 6× accelerations with only 4 or 8 calibration lines in 2D Cartesian acquisitions, corresponding to 22.0% and 19.5% undersampling rates. The high-quality results and scanning specificity make the proposed method hold the potential for further accelerating the data acquisition of parallel MRI.

On retrospective k-space subsampling schemes for deep MRI reconstruction

Acquiring fully-sampled MRI k-space data is time-consuming, and collecting accelerated data can reduce the acquisition time. Employing 2D Cartesian-rectilinear subsampling schemes is a conventional approach for accelerated acquisitions; however, this often results in imprecise reconstructions, even with the use of Deep Learning (DL), especially at high acceleration factors. Non-rectilinear or non-Cartesian trajectories can be implemented in MRI scanners as alternative subsampling options. This work investigates the impact of the k-space subsampling scheme on the quality of reconstructed accelerated MRI measurements produced by trained DL models. The Recurrent Variational Network (RecurrentVarNet) was used as the DL-based MRI-reconstruction architecture. Cartesian, fully-sampled multi-coil k-space measurements from three datasets were retrospectively subsampled with different accelerations using eight distinct subsampling schemes: four Cartesian-rectilinear, two Cartesian non-rectilinear, and two non-Cartesian. Experiments were conducted in two frameworks: scheme-specific, where a distinct model was trained and evaluated for each dataset-subsampling scheme pair, and multi-scheme, where for each dataset a single model was trained on data randomly subsampled by any of the eight schemes and evaluated on data subsampled by all schemes. In both frameworks, RecurrentVarNets trained and evaluated on non-rectilinearly subsampled data demonstrated superior performance, particularly for high accelerations. In the multi-scheme setting, reconstruction performance on rectilinearly subsampled data improved when compared to the scheme-specific experiments. Our findings demonstrate the potential for using DL-based methods, trained on non-rectilinearly subsampled measurements, to optimize scan time and image quality.

MA-RECON: Mask-aware deep-neural-network for robust fast MRI k-space interpolation

BACKGROUND AND OBJECTIVE

High-quality reconstruction of MRI images from under-sampled 'k-space' data, which is in the Fourier domain, is crucial for shortening MRI acquisition times and ensuring superior temporal resolution. Over recent years, a wealth of deep neural network (DNN) methods have emerged, aiming to tackle the complex, ill-posed inverse problem linked to this process. However, their instability against variations in the acquisition process and anatomical distribution exposes a deficiency in the generalization of relevant physical models within these DNN architectures. The goal of our work is to enhance the generalization capabilities of DNN methods for k-space interpolation by introducing 'MA-RECON', an innovative mask-aware DNN architecture and associated training method.

METHODS

Unlike preceding approaches, our 'MA-RECON' architecture encodes not only the observed data but also the under-sampling mask within the model structure. It implements a tailored training approach that leverages data generated with a variety of under-sampling masks to stimulate the model's generalization of the under-sampled MRI reconstruction problem. Therefore, effectively represents the associated inverse problem, akin to the classical compressed sensing approach.

RESULTS

The benefits of our MA-RECON approach were affirmed through rigorous testing with the widely accessible fastMRI dataset. Compared to standard DNN methods and DNNs trained with under-sampling mask augmentation, our approach demonstrated superior generalization capabilities. This resulted in a considerable improvement in robustness against variations in both the acquisition process and anatomical distribution, especially in regions with pathology.

CONCLUSION

In conclusion, our mask-aware strategy holds promise for enhancing the generalization capacity and robustness of DNN-based methodologies for MRI reconstruction from undersampled k-space data. Code is available in the following link: https://github.com/nitzanavidan/PD_Recon.

Semiautomated Segmentation and Analysis of Airway Lumen in Pediatric Patients Using Ultra Short Echo Time MRI

RATIONALE AND OBJECTIVES

Ultra short echo time (UTE) magnetic resonance imaging (MRI) pulse sequences have shown promise for airway assessment, but the feasibility and repeatability in the pediatric lung are unknown. The purpose of this work was to develop a semiautomated UTE MRI airway segmentation pipeline from the trachea-to-tertiary airways in pediatric participants and assess repeatability and lumen diameter correlations to lung function.

MATERIALS AND METHODS

A total of 29 participants (n = 7 healthy, n = 11 cystic fibrosis, n = 6 asthma, and n = 5 ex-preterm), aged 7-18 years, were imaged using a 3D stack-of-spirals UTE examination at 3 T. Two independent observers performed airway segmentations using a pipeline developed in-house; observer 1 repeated segmentations 1 month later. Segmentations were extracted using region-growing with leak detection, then manually edited if required. The airway trees were skeletonized, pruned, and labeled. Airway lumen diameter measurements were extracted using ray casting. Intra- and interobserver variability was assessed using the Sørensen-Dice coefficient (DSC) and intra-class correlation coefficient (ICC). Correlations between lumen diameter and pulmonary function were assessed using Spearman's correlation coefficient.

RESULTS

For airway segmentations and lumen diameter, intra- and interobserver DSCs were 0.88 and 0.80, while ICCs were 0.95 and 0.89, respectively. The variability increased from the trachea-to-tertiary airways for intra- (DSC: 0.91-0.64; ICC: 0.91-0.49) and interobserver (DSC: 0.84-0.51; ICC: 0.89-0.21) measurements. Lumen diameter was significantly correlated with forced expiratory volume in 1 second and forced vital capacity (P < .05).

CONCLUSION

UTE MRI airway segmentation from the trachea-to-tertiary airways in pediatric participants across a range of diseases is feasible. The UTE MRI-derived lumen measurements were repeatable and correlated with lung function.

A Physics-Based Algorithm to Universally Standardize Routinely Obtained Clinical T2-Weighted Images

RATIONALE AND OBJECTIVES

MR images can be challenging for machine learning and other large-scale analyses because most clinical images, for example, T2-weighted (T2w) images, reflect not only the biologically relevant T2 of tissue but also hardware and acquisition parameters that vary from site to site. Quantitative T2 mapping avoids these confounds because it quantitatively isolates the biological parameter of interest, thus representing a universal standardization across sites. However, efforts to incorporate quantitative mapping sequences into routine clinical practice have seen slow adoption. Here we show, for the first time, that the routine T2w complex raw dataset can be successfully regarded as a quantitative mapping sequence that can be reconstructed with classical optimization methods and physics-based constraints.

MATERIALS AND METHODS

While previous constrained reconstruction methods are unable to reconstruct a T2 map based on this data, the expanding-constrained alternating minimization for parameter mapping (e-CAMP), which employs stepwise initialization, a linearized version of the exponential model and a phase conjugacy constraint, is demonstrated to provide useful quantitative maps directly from a vendor T2w single image data.

RESULTS

This paper introduces the method and demonstrates its performance using simulations, retrospectively undersampled brain images, and prospectively acquired T2w images taken on both phantom and brain.

CONCLUSION

Because T2w scans are included in nearly every protocol, this approach could open the door to creating large, standardized datasets without requiring widespread changes in clinical protocols.

Magnetic resonance coherence pathway unraveling

Efficiently acquiring multi-contrast magnetic resonance imaging data is crucial for patient comfort and clinical throughput. Developing scan acceleration methods tailored for specific applications drastically improves the value of an MRI examination. Here, we propose a novel method to control the aliasing of simultaneously acquired images of multiple spin echo coherence pathways with the goal of producing high quality multi-contrast images from a single acquisition. Modulating the radiofrequency phase of several pulses applied in brief succession also uniquely modulates the phase of spin echo coherence pathways. A method, termed magnetic resonance coherence pathway unraveling (MR-CPU), to control the aliasing of simultaneously acquired coherence pathway images is developed here along with parallel imaging-based reconstruction methods to separate them. MR-CPU was validated in phantom experiments and tested in vivo. High levels of correlation between reference pathway images and MR-CPU-derived coherence pathway images were found from the phantom experiments. Minimal artifacts arising from the separation of the overlapped coherence pathway images were observed in vivo. MR-CPU provides a novel mechanism through which to acquire and separate multiple overlapped coherence pathway images, thus adding to the diagnostic potential of an MRI exam without the penalty of additional scan time.

A characterization of cardiac-induced noise in R2 * maps of the brain

PURPOSE

Cardiac pulsation increases the noise level in brain maps of the transverse relaxation rate R2 *. Cardiac-induced noise is challenging to mitigate during the acquisition of R2 * mapping data because its characteristics are unknown. In this work, we aim to characterize cardiac-induced noise in brain maps of the MRI parameter R2 *.

METHODS

We designed a sampling strategy to acquire multi-echo 3D data in 12 intervals of the cardiac cycle, monitored with a fingertip pulse-oximeter. We measured the amplitude of cardiac-induced noise in this data and assessed the effect of cardiac pulsation on R2 * maps computed across echoes. The area of k-space that contains most of the cardiac-induced noise in R2 * maps was then identified. Based on these characteristics, we introduced a tentative sampling strategy that aims to mitigate cardiac-induced noise in R2 * maps of the brain.

RESULTS

In inferior brain regions, cardiac pulsation accounts for R2 * variations of up to 3 s-1 across the cardiac cycle (i.e., ∼35% of the overall variability). Cardiac-induced fluctuations occur throughout the cardiac cycle, with a reduced intensity during the first quarter of the cycle. A total of 50% to 60% of the overall cardiac-induced noise is localized near the k-space center (k < 0.074 mm-1 ). The tentative cardiac noise mitigation strategy reduced the variability of R2 * maps across repetitions by 11% in the brainstem and 6% across the whole brain.

CONCLUSION

We provide a characterization of cardiac-induced noise in brain R2 * maps that can be used as a basis for the design of mitigation strategies during data acquisition.

AliasNet: Alias artefact suppression network for accelerated phase-encode MRI

Sparse reconstruction is an important aspect of MRI, helping to reduce acquisition time and improve spatial-temporal resolution. Popular methods are based mostly on compressed sensing (CS), which relies on the random sampling of k-space to produce incoherent (noise-like) artefacts. Due to hardware constraints, 1D Cartesian phase-encode under-sampling schemes are popular for 2D CS-MRI. However, 1D under-sampling limits 2D incoherence between measurements, yielding structured aliasing artefacts (ghosts) that may be difficult to remove assuming a 2D sparsity model. Reconstruction algorithms typically deploy direction-insensitive 2D regularisation for these direction-associated artefacts. Recognising that phase-encode artefacts can be separated into contiguous 1D signals, we develop two decoupling techniques that enable explicit 1D regularisation and leverage the excellent 1D incoherence characteristics. We also derive a combined 1D + 2D reconstruction technique that takes advantage of spatial relationships within the image. Experiments conducted on retrospectively under-sampled brain and knee data demonstrate that combination of the proposed 1D AliasNet modules with existing 2D deep learned (DL) recovery techniques leads to an improvement in image quality. We also find AliasNet enables a superior scaling of performance compared to increasing the size of the original 2D network layers. AliasNet therefore improves the regularisation of aliasing artefacts arising from phase-encode under-sampling, by tailoring the network architecture to account for their expected appearance. The proposed 1D + 2D approach is compatible with any existing 2D DL recovery technique deployed for this application.

Rapid MR elastography of the liver for subsecond stiffness sampling

PURPOSE

Depicting the stiffness of biological soft tissues, MR elastography (MRE) has a wide range of diagnostic applications. The purpose of this study was to improve the temporal resolution of 2D hepatic MRE in order to provide more rapid feedback on the quality of the wavefield and ensure better temporal sampling of respiration-induced stiffness changes.

METHODS

We developed a rapid MRE sequence that uses 2D segmented gradient-echo spiral readout to encode 40 Hz harmonic vibrations and generate stiffness maps within 625 ms. We demonstrate the use of this technique as a rapid test for shear wave amplitudes and overall MRE image quality and as a method for monitoring respiration-induced stiffness changes in the liver in comparison to 3D MRE and ultrasound-based time-harmonic elastography.

RESULTS

Subsecond MRE allowed monitoring of increasing shear wave amplitudes in the liver with increasing levels of external stimulation within a single breath-hold. Furthermore, the technique detected respiration-induced changes in liver stiffness with peak values (1.83 ± 0.22 m/s) at end-inspiration, followed by softer values during forced abdominal pressure (1.60 ± 0.22 m/s) and end-expiration (1.49 ± 0.22 m/s). The effects of inspiration and expiration were confirmed by time-harmonic elastography.

CONCLUSION

Our results suggest that subsecond MRE of the liver is useful for checking MRE driver settings and monitoring breathing-induced changes in liver stiffness in near real time.

Accelerated submillimeter wave-encoded magnetic resonance imaging via deep untrained neural network

BACKGROUND

Wave gradient encoding can adequately utilize coil sensitivity profiles to facilitate higher accelerations in parallel magnetic resonance imaging (pMRI). However, there are limitations in mainstream pMRI and a few deep learning (DL) methods for recovering missing data under wave encoding framework: the former is prone to introduce errors from the auto-calibration signals (ACS) signal acquisition and is time-consuming, while the latter requires a large amount of training data.

PURPOSE

To tackle the above issues, an untrained neural network (UNN) model incorporating wave-encoded physical properties and deep generative model, named WDGM, was proposed with additional ACS- and training data-free.

METHODS

Generally, the proposed method can provide powerful missing data interpolation capability using the wave physical encoding framework and designed UNN to characterize the MR image (k-space data) priors. Specifically, the MRI reconstruction combining physical wave encoding and elaborate UNN is modeled as a generalized minimization problem. The designation of UNN is driven by the coil sensitivity maps (CSM) smoothness and k-space linear predictability. And then, the iterative paradigm to recover the full k-space signal is determined by the projected gradient descent, and the complex computation is unrolled to the network with optimized parameters by the optimizer. Simulated wave encoding and in vivo experiments are exploited to demonstrate the feasibility of the proposed method. The best quantitative metrics RMSE/SSIM/PSNR of 0.0413, 0.9514, and 37.4862 gave competitive results in all experiments with at least six-fold acceleration, respectively.

RESULTS

In vivo experiments of human brains and knees showed that the proposed method can achieve comparable reconstruction quality and even has superiority relative to the comparison, especially at a high resolution of 0.67 mm and fewer ACS. In addition, the proposed method has a higher computational efficiency achieving a computation time of 9.6 s/per slice.

CONCLUSIONS

The model proposed in this work addresses two limitations of MRI reconstruction in the wave encoding framework. The first is to eliminate the need for ACS signal acquisition to perform the time-consuming calibration process and to avoid errors such as motion during the acquisition procedure. Furthermore, the proposed method has clinical application friendly without the need to prepare large training datasets, which is difficult in the clinical. All results of the proposed method demonstrate more confidence in both quantitative and qualitative metrics. In addition, the proposed method can achieve higher computational efficiency.

Parallel-stream fusion of scan-specific and scan-general priors for learning deep MRI reconstruction in low-data regimes

Magnetic resonance imaging (MRI) is an essential diagnostic tool that suffers from prolonged scan times. Reconstruction methods can alleviate this limitation by recovering clinically usable images from accelerated acquisitions. In particular, learning-based methods promise performance leaps by employing deep neural networks as data-driven priors. A powerful approach uses scan-specific (SS) priors that leverage information regarding the underlying physical signal model for reconstruction. SS priors are learned on each individual test scan without the need for a training dataset, albeit they suffer from computationally burdening inference with nonlinear networks. An alternative approach uses scan-general (SG) priors that instead leverage information regarding the latent features of MRI images for reconstruction. SG priors are frozen at test time for efficiency, albeit they require learning from a large training dataset. Here, we introduce a novel parallel-stream fusion model (PSFNet) that synergistically fuses SS and SG priors for performant MRI reconstruction in low-data regimes, while maintaining competitive inference times to SG methods. PSFNet implements its SG prior based on a nonlinear network, yet it forms its SS prior based on a linear network to maintain efficiency. A pervasive framework for combining multiple priors in MRI reconstruction is algorithmic unrolling that uses serially alternated projections, causing error propagation under low-data regimes. To alleviate error propagation, PSFNet combines its SS and SG priors via a novel parallel-stream architecture with learnable fusion parameters. Demonstrations are performed on multi-coil brain MRI for varying amounts of training data. PSFNet outperforms SG methods in low-data regimes, and surpasses SS methods with few tens of training samples. On average across tasks, PSFNet achieves 3.1 dB higher PSNR, 2.8% higher SSIM, and 0.3 × lower RMSE than baselines. Furthermore, in both supervised and unsupervised setups, PSFNet requires an order of magnitude lower samples compared to SG methods, and enables an order of magnitude faster inference compared to SS methods. Thus, the proposed model improves deep MRI reconstruction with elevated learning and computational efficiency.

Equilibrated Zeroth-Order Unrolled Deep Network for Parallel MR Imaging

In recent times, model-driven deep learning has evolved an iterative algorithm into a cascade network by replacing the regularizer's first-order information, such as the (sub)gradient or proximal operator, with a network module. This approach offers greater explainability and predictability compared to typical data-driven networks. However, in theory, there is no assurance that a functional regularizer exists whose first-order information matches the substituted network module. This implies that the unrolled network output may not align with the regularization models. Furthermore, there are few established theories that guarantee global convergence and robustness (regularity) of unrolled networks under practical assumptions. To address this gap, we propose a safeguarded methodology for network unrolling. Specifically, for parallel MR imaging, we unroll a zeroth-order algorithm, where the network module serves as a regularizer itself, allowing the network output to be covered by a regularization model. Additionally, inspired by deep equilibrium models, we conduct the unrolled network before backpropagation to converge to a fixed point and then demonstrate that it can tightly approximate the actual MR image. We also prove that the proposed network is robust against noisy interferences if the measurement data contain noise. Finally, numerical experiments indicate that the proposed network consistently outperforms state-of-the-art MRI reconstruction methods, including traditional regularization and unrolled deep learning techniques.

Six-minute, in vivo MRI quantification of proximal femur trabecular bone 3D microstructure

BACKGROUND

Assessment of proximal femur trabecular bone microstructure in vivo by magnetic resonance imaging has recently been validated for acquiring information independent of bone mineral density in osteoporotic patients. However, the requisite signal-to-noise ratio (SNR) and resolution for interrogation of the trabecular microstructure at this anatomical location prolongs the scan duration and renders the imaging protocol clinically infeasible. Parallel imaging and compressed sensing (PICS) techniques can reduce the scan duration of the imaging protocol without substantially compromising image quality. The present work investigates the limits of acceleration for a commonly used PICS technique, ℓ1-ESPIRiT, for the purpose of quantifying measures of trabecular bone microarchitecture. Based on a desired error tolerance, a six-minute, prospectively accelerated variant of the imaging protocol was developed and assessed for intersession reproducibility and agreement with the longer reference scan.

PURPOSE

To investigate the limits of acceleration for MRI-based trabecular bone quantification by parallel imaging and compressed sensing reconstruction, and to develop a prototypical imaging protocol for assessing the proximal femur microstructure in a clinically practical scan time.

METHODS

Healthy participants (n = 11) were scanned by a 3D balanced steady-state free precession (bSSFP) sequence satisfying the Nyquist criterion with a scan duration of about 18 min. The raw data were retrospectively undersampled and reconstructed to mimic various acceleration factors ranging from 2 to 6. Trabecular volumes-of-interest in four major femoral regions (greater trochanter, intertrochanteric region, femoral neck, and femoral head) were analyzed and six relevant measures of trabecular bone microarchitecture (bone volume fraction, surface-to-curve ratio, erosion index, elastic modulus, trabecular thickness, plates-to-rods ratio) were obtained for images of all accelerations. To assess agreement, median percent error and intraclass correlation coefficients (ICCs) were computed using the fully-sampled data as reference. Based on this analysis, a prospectively 3-fold accelerated sequence with a duration of about 6 min was developed and the analysis was repeated.

RESULTS

A prospective acceleration factor of 3 demonstrated comparable performance in reproducibility and absolute agreement to the fully-sampled scan. The median CoV over all image-derived metrics was generally <6 % and ICCs >0.70. Also, measurements from prospectively 3-fold accelerated scans demonstrated in general median percent errors of <7 % and ICCs >0.70.

CONCLUSION

The present work proposes a method to make in vivo quantitative assessment of proximal femur trabecular microstructure with a clinically practical scan duration of about 6 min.

AI-assisted compressed sensing and parallel imaging sequences for MRI of patients with nasopharyngeal carcinoma: comparison of their capabilities in terms of examination time and image quality

OBJECTIVE

To compare examination time and image quality between artificial intelligence (AI)-assisted compressed sensing (ACS) technique and parallel imaging (PI) technique in MRI of patients with nasopharyngeal carcinoma (NPC).

METHODS

Sixty-six patients with pathologically confirmed NPC underwent nasopharynx and neck examination using a 3.0-T MRI system. Transverse T2-weighted fast spin-echo (FSE) sequence, transverse T1-weighted FSE sequence, post-contrast transverse T1-weighted FSE sequence, and post-contrast coronal T1-weighted FSE were obtained by both ACS and PI techniques, respectively. The signal-to-noise ratio (SNR), contrast-to-noise ratio (CNR), and duration of scanning of both sets of images analyzed by ACS and PI techniques were compared. The images from the ACS and PI techniques were scored for lesion detection, margin sharpness of lesions, artifacts, and overall image quality using the 5-point Likert scale.

RESULTS

The examination time with ACS technique was significantly shorter than that with PI technique (p < 0.0001). The comparison of SNR and CNR showed that ACS technique was significantly superior with PI technique (p < 0.005). Qualitative image analysis showed that the scores of lesion detection, margin sharpness of lesions, artifacts, and overall image quality were higher in the ACS sequences than those in the PI sequences (p < 0.0001). Inter-observer agreement was evaluated for all qualitative indicators for each method, in which the results showed satisfactory-to-excellent agreement (p < 0.0001).

CONCLUSION

Compared with the PI technique, the ACS technique for MR examination of NPC can not only shorten scanning time but also improve image quality.

CLINICAL RELEVANCE STATEMENT

The artificial intelligence (AI)-assisted compressed sensing (ACS) technique shortens examination time for patients with nasopharyngeal carcinoma, while improving the image quality and examination success rate, which will benefit more patients.

KEY POINTS

• Compared with the parallel imaging (PI) technique, the artificial intelligence (AI)-assisted compressed sensing (ACS) technique not only reduced examination time, but also improved image quality. • Artificial intelligence (AI)-assisted compressed sensing (ACS) pulls the state-of-the-art deep learning technique into the reconstruction procedure and helps find an optimal balance of imaging speed and image quality.

WKGM: weighted k-space generative model for parallel imaging reconstruction

Deep learning based parallel imaging (PI) has made great progress in recent years to accelerate MRI. Nevertheless, it still has some limitations: for example, the robustness and flexibility of existing methods are greatly deficient. In this work, we propose a method to explore the k-space domain learning via robust generative modeling for flexible calibrationless PI reconstruction, coined the weighted k-space generative model (WKGM). Specifically, WKGM is a generalized k-space domain model, where the k-space weighting technology and high-dimensional space augmentation design are efficiently incorporated for score-based generative model training, resulting in good and robust reconstructions. In addition, WKGM is flexible and thus can be synergistically combined with various traditional k-space PI models, which can make full use of the correlation between multi-coil data and realize calibrationless PI. Even though our model was trained on only 500 images, experimental results with varying sampling patterns and acceleration factors demonstrate that WKGM can attain state-of-the-art reconstruction results with the well learned k-space generative prior.

Sampling strategies and integrated reconstruction for reducing distortion and boundary slice aliasing in high-resolution 3D diffusion MRI

PURPOSE

To develop a new method for high-fidelity, high-resolution 3D multi-slab diffusion MRI with minimal distortion and boundary slice aliasing.

METHODS

Our method modifies 3D multi-slab imaging to integrate blip-reversed acquisitions for distortion correction and oversampling in the slice direction (kz ) for reducing boundary slice aliasing. Our aim is to achieve robust acceleration to keep the scan time the same as conventional 3D multi-slab acquisitions, in which data are acquired with a single direction of blip traversal and without kz -oversampling. We employ a two-stage reconstruction. In the first stage, the blip-up/down images are respectively reconstructed and analyzed to produce a field map for each diffusion direction. In the second stage, the blip-reversed data and the field map are incorporated into a joint reconstruction to produce images that are corrected for distortion and boundary slice aliasing.

RESULTS

We conducted experiments at 7T in six healthy subjects. Stage 1 reconstruction produces images from highly under-sampled data (R = 7.2) with sufficient quality to provide accurate field map estimation. Stage 2 joint reconstruction substantially reduces distortion artifacts with comparable quality to fully-sampled blip-reversed results (2.4× scan time). Whole-brain in-vivo results acquired at 1.22 mm and 1.05 mm isotropic resolutions demonstrate improved anatomical fidelity compared to conventional 3D multi-slab imaging. Data demonstrate good reliability and reproducibility of the proposed method over multiple subjects.

CONCLUSION

The proposed acquisition and reconstruction framework provide major reductions in distortion and boundary slice aliasing for 3D multi-slab diffusion MRI without increasing the scan time, which can potentially produce high-quality, high-resolution diffusion MRI.

3D FRONSAC with PSF reconstruction

PURPOSE

This study extends the Fast ROtary Nonlinear Spatial ACquisition (FRONSAC) method to include 3D acquisitions and reconstructions. It uses a transform domain reconstruction which is needed to make 3D reconstructions practical and provides new insights into how parallel imaging performance is enhanced by FRONSAC encoding.

METHODS

This work developed the first examples of FRONSAC incorporated into a 3D acquisition. 3D FRONSAC was tested on human subjects with both simple gradient echo and MPRAGE Cartesian acquisitions. The quality of the 3D FRONSAC images was evaluated using structural similarity index measure (SSIM), and normalized root mean square error (NRMSE).

RESULTS

FRONSAC encoding did not significantly modify the contrast obtained in either sequence, but it substantially improves the image quality of undersampled reconstruction. FRONSAC images have reduced undersampling ghosts and consistently improved SSIM and NRMSE.

CONCLUSIONS

Acquisition and reconstruction of 3D FRONSAC images are feasible, and the additional FRONSAC encoding improves image quality in highly undersampled images.

Joint Image Reconstruction and Super-Resolution for Accelerated Magnetic Resonance Imaging

Magnetic resonance (MR) image reconstruction and super-resolution are two prominent techniques to restore high-quality images from undersampled or low-resolution k-space data to accelerate MR imaging. Combining undersampled and low-resolution acquisition can further improve the acceleration factor. Existing methods often treat the techniques of image reconstruction and super-resolution separately or combine them sequentially for image recovery, which can result in error propagation and suboptimal results. In this work, we propose a novel framework for joint image reconstruction and super-resolution, aiming to efficiently image recovery and enable fast imaging. Specifically, we designed a framework with a reconstruction module and a super-resolution module to formulate multi-task learning. The reconstruction module utilizes a model-based optimization approach, ensuring data fidelity with the acquired k-space data. Moreover, a deep spatial feature transform is employed to enhance the information transition between the two modules, facilitating better integration of image reconstruction and super-resolution. Experimental evaluations on two datasets demonstrate that our proposed method can provide superior performance both quantitatively and qualitatively.

PARCEL: Physics-Based Unsupervised Contrastive Representation Learning for Multi-Coil MR Imaging

With the successful application of deep learning to magnetic resonance (MR) imaging, parallel imaging techniques based on neural networks have attracted wide attention. However, in the absence of high-quality, fully sampled datasets for training, the performance of these methods is limited. And the interpretability of models is not strong enough. To tackle this issue, this paper proposes a Physics-bAsed unsupeRvised Contrastive rEpresentation Learning (PARCEL) method to speed up parallel MR imaging. Specifically, PARCEL has a parallel framework to contrastively learn two branches of model-based unrolling networks from augmented undersampled multi-coil k-space data. A sophisticated co-training loss with three essential components has been designed to guide the two networks in capturing the inherent features and representations for MR images. And the final MR image is reconstructed with the trained contrastive networks. PARCEL was evaluated on two vivo datasets and compared to five state-of-the-art methods. The results show that PARCEL is able to learn essential representations for accurate MR reconstruction without relying on fully sampled datasets. The code will be made available at https://github.com/ternencewu123/PARCEL.

Emerging Trends in Fast MRI Using Deep-Learning Reconstruction on Undersampled k-Space Data: A Systematic Review

Magnetic Resonance Imaging (MRI) is an essential medical imaging modality that provides excellent soft-tissue contrast and high-resolution images of the human body, allowing us to understand detailed information on morphology, structural integrity, and physiologic processes. However, MRI exams usually require lengthy acquisition times. Methods such as parallel MRI and Compressive Sensing (CS) have significantly reduced the MRI acquisition time by acquiring less data through undersampling k-space. The state-of-the-art of fast MRI has recently been redefined by integrating Deep Learning (DL) models with these undersampled approaches. This Systematic Literature Review (SLR) comprehensively analyzes deep MRI reconstruction models, emphasizing the key elements of recently proposed methods and highlighting their strengths and weaknesses. This SLR involves searching and selecting relevant studies from various databases, including Web of Science and Scopus, followed by a rigorous screening and data extraction process using the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. It focuses on various techniques, such as residual learning, image representation using encoders and decoders, data-consistency layers, unrolled networks, learned activations, attention modules, plug-and-play priors, diffusion models, and Bayesian methods. This SLR also discusses the use of loss functions and training with adversarial networks to enhance deep MRI reconstruction methods. Moreover, we explore various MRI reconstruction applications, including non-Cartesian reconstruction, super-resolution, dynamic MRI, joint learning of reconstruction with coil sensitivity and sampling, quantitative mapping, and MR fingerprinting. This paper also addresses research questions, provides insights for future directions, and emphasizes robust generalization and artifact handling. Therefore, this SLR serves as a valuable resource for advancing fast MRI, guiding research and development efforts of MRI reconstruction for better image quality and faster data acquisition.

[Physical model-based cascaded generative adversarial networks for accelerating quantitative multi-parametric magnetic resonance imaging]

OBJECTIVE

To explore the feasibility and interpretation of physical model- based cascaded generative adversarial networks for accelerating quantitative multi-echo multi-parametric magnetic resonance imaging using raw multi-echo multicoil k-space data.

METHODS

A physical model-based cascaded generative adversarial network is proposed to enhance image feature information to obtain high-quality reconstructed images using joint training of multi-domain information and learning of key parameters required for image reconstruction through a system matrix and adaptively optimizing the k-space generator and image generator structures. Raw multi-echo multi-coil k-space data are used to accelerate multi-contrast multi-parametric magnetic resonance imaging. A physically driven deep learning reconstruction method is used to increase the generalization capability and improve the model performance by building a system matrix function instead of direct end-to-end training of the model.

RESULTS

In terms of overall image quality, the proposed model achieved significant improvements compared to other methods. On an 80- case test set, the average PSNR value of the reconstructed images was 34.13, SSIM was 0.965, and NRMSE was 0.114. In terms of multi-contrast multi-parametric image reconstruction, the model achieved PSNR values of 38.87 for PDW, 35.62 for T1W, and 34.38 for T2* Map, which were significantly better than those of other methods for quantitative evaluation. The model also produced clearer features of the brain gray matter, white matter, and cerebrospinal fluid. Furthermore, compared with the existing methods with a reconstruction time difference of less than 10%, the proposed method achieved the highest improvement of up to 20% in the metrics of PSNR, SSIM, and NRMSE.

CONCLUSION

Compared with other existing methods, the physical model-based cascaded generative adversarial networks can reconstruct more image details and features, thus improving the quality and accuracy of the reconstructed images.

Highly accelerated intracranial time-of-flight magnetic resonance angiography using wave-encoding

PURPOSE

To develop an accelerated 3D intracranial time-of-flight (TOF) magnetic resonance angiography (MRA) sequence with wave-encoding (referred to as 3D wave-TOF) and to evaluate two variants: wave-controlled aliasing in parallel imaging (CAIPI) and compressed-sensing wave (CS-wave).

METHODS

A wave-TOF sequence was implemented on a 3 T clinical scanner. Wave-encoded and Cartesian k-space datasets from six healthy volunteers were retrospectively and prospectively undersampled with 2D-CAIPI sampling and variable-density Poisson disk sampling. 2D-CAIPI, wave-CAIPI, standard CS, and CS-wave schemes were compared at various acceleration factors. Flow-related artifacts in wave-TOF were investigated, and a set of practicable wave parameters was developed. Quantitative analysis of wave-TOF and traditional Cartesian TOF MRA was performed by comparing the contrast-to-background ratio between the vessel and background tissue in source images, and the structural similarity index measure (SSIM) between the maximum intensity projection images from accelerated acquisitions and their respective fully sampled references.

RESULTS

Flow-related artifacts caused by the wave-encoding gradients in wave-TOF were eliminated by properly chosen parameters. Images from wave-CAIPI and CS-wave acquisitions had a higher SNR and better-preserved contrast than traditional parallel imaging (PI) and CS methods. Maximum intensity projection images from wave-CAIPI and CS-wave acquisitions had a cleaner background, with vessels that were better depicted. Quantitative analyses indicated that wave-CAIPI had the highest contrast-to-background ratio, SSIM, and vessel-masked SSIM among the sampling schemes studied, followed by the CS-wave acquisition.

CONCLUSION

3D wave-TOF improves the capability of accelerated MRA and provides better image quality at higher acceleration factors compared to traditional PI- or CS-accelerated TOF, suggesting the potential use of wave-TOF in cerebrovascular disease.

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.

Spatially and velocity-selective magnetization preparation for noncontrast-enhanced peripheral MR angiography

The purpose of the current study was to develop spatially and velocity-selective (SVS) magnetization preparation pulses for noncontrast-enhanced peripheral MR angiography (MRA) to provide comparisons with velocity-selective (VS) MRA with comparison to velocity-selective (VS). VS preparation pulses were designed by concatenating multiple excitation steps, each of which was a combination of a hard RF pulse, VS unipolar gradient pulses, and refocusing RF pulses. SVS preparation pulses were designed by replacing the hard RF pulse with a sinc-shaped RF pulse combined with a symmetric tripolar gradient pulse (which does not perturb the velocity encoding by the VS unipolar gradient pulses). Numerical simulations were performed to verify the intended hybrid excitation selectivity of SVS pulses taking account of tissue relaxation, magnetic field errors, and eddy currents. In vivo experiments were performed in healthy subjects to verify the hybrid excitation selectivity, as well as to demonstrate the visualization of the entire peripheral arteries using six-station protocols. As demonstrated by numerical simulations, SVS preparation yielded a notch-shaped longitudinal magnetization (M_z )-velocity response within the spatial stopband (the same as VS preparation) and preserved the M_z of spins outside the stopband, regardless of its velocity. We confirmed these observations also through in vivo tests with good agreement in normalized arterial and muscle signal intensities. In six-station peripheral MRA experiments, the proposed SVS-MRA yielded significantly higher arterial signal-to-noise ratio (SNR) (51.6 ± 14.3 vs. 38.9 ± 10.9; p < 0.001) and contrast-to-noise ratio (CNR) (41.2 ± 13.0 vs. 31.3 ± 10.5; p < 0.001) compared with VS-MRA. The proposed SVS-MRA improves arterial SNR and CNR compared with VS-MRA by mitigating undesired presaturation of arterial blood upstream the imaging field of view.

Accelerating 4D image reconstruction for magnetic resonance-guided radiotherapy

Background and purpose

Physiological motion impacts the dose delivered to tumours and vital organs in external beam radiotherapy and particularly in particle therapy. The excellent soft-tissue demarcation of 4D magnetic resonance imaging (4D-MRI) could inform on intra-fractional motion, but long image reconstruction times hinder its use in online treatment adaptation. Here we employ techniques from high-performance computing to reduce 4D-MRI reconstruction times below two minutes to facilitate their use in MR-guided radiotherapy.

Material and methods

Four patients with pancreatic adenocarcinoma were scanned with a radial stack-of-stars gradient echo sequence on a 1.5T MR-Linac. Fast parallelised open-source implementations of the extra-dimensional golden-angle radial sparse parallel algorithm were developed for central processing unit (CPU) and graphics processing unit (GPU) architectures. We assessed the impact of architecture, oversampling and respiratory binning strategy on 4D-MRI reconstruction time and compared images using the structural similarity (SSIM) index against a MATLAB reference implementation. Scaling and bottlenecks for the different architectures were studied using multi-GPU systems.

Results

All reconstructed 4D-MRI were identical to the reference implementation (SSIM > 0.99). Images reconstructed with overlapping respiratory bins were sharper at the cost of longer reconstruction times. The CPU  + GPU implementation was over 17 times faster than the reference implementation, reconstructing images in 60 ± 1 s and hyper-scaled using multiple GPUs.

Conclusion

Respiratory-resolved 4D-MRI reconstruction times can be reduced using high-performance computing methods for online workflows in MR-guided radiotherapy with potential applications in particle therapy.

K-UNN: k-space interpolation with untrained neural network

Recently, untrained neural networks (UNNs) have shown satisfactory performances for MR image reconstruction on random sampling trajectories without using additional full-sampled training data. However, the existing UNN-based approaches lack the modeling of physical priors, resulting in poor performance in some common scenarios (e.g., partial Fourier (PF), regular sampling, etc.) and the lack of theoretical guarantees for reconstruction accuracy. To bridge this gap, we propose a safeguarded k-space interpolation method for MRI using a specially designed UNN with a tripled architecture driven by three physical priors of the MR images (or k-space data), including transform sparsity, coil sensitivity smoothness, and phase smoothness. We also prove that the proposed method guarantees tight bounds for interpolated k-space data accuracy. Finally, ablation experiments show that the proposed method can characterize the physical priors of MR images well. Additionally, experiments show that the proposed method consistently outperforms traditional parallel imaging methods and existing UNNs, and is even competitive against supervised-trained deep learning methods in PF and regular undersampling reconstruction.

Acquisition sequences and reconstruction methods for fast chemical exchange saturation transfer imaging

Chemical exchange saturation transfer (CEST) imaging is an emerging molecular magnetic resonance imaging (MRI) technique that has been developed and employed in numerous diseases. Based on the unique saturation transfer principle, a family of CEST-detectable biomolecules in vivo have been found capable of providing valuable diagnostic information. However, CEST MRI needs a relatively long scan time due to the common long saturation labeling module and typical acquisition of multiple frequency offsets and signal averages, limiting its widespread clinical applications. So far, a plethora of imaging schemes and techniques has been developed to accelerate CEST MRI. In this review, the key acquisition and reconstruction methods for fast CEST imaging are summarized from a practical and systematic point of view. The first acquisition sequence section describes the major development of saturation schemes, readout patterns, ultrafast z-spectroscopy, and saturation-editing techniques for rapid CEST imaging. The second reconstruction method section lists the important advances of parallel imaging, compressed sensing, sparsity in the z-spectrum, and algorithms beyond the Fourier transform for speeding up CEST MRI.

Clinical Impact of Deep Learning Reconstruction in MRI

Deep learning has been recognized as a paradigm-shifting tool in radiology. Deep learning reconstruction (DLR) has recently emerged as a technology used in the image reconstruction process of MRI, which is an essential procedure in generating MR images. Denoising, which is the first DLR application to be realized in commercial MRI scanners, improves signal-to-noise ratio. When applied to lower magnetic field-strength scanners, the signal-to-noise ratio can be increased without extending the imaging time, and image quality is comparable to that of higher-field-strength scanners. Shorter imaging times decrease patient discomfort and reduce MRI scanner running costs. The incorporation of DLR into accelerated acquisition imaging techniques, such as parallel imaging or compressed sensing, shortens the reconstruction time. DLR is based on supervised learning using convolutional layers and is divided into the following three categories: image domain, k-space learning, and direct mapping types. Various studies have reported other derivatives of DLR, and several have shown the feasibility of DLR in clinical practice. Although DLR efficiently reduces Gaussian noise from MR images, denoising makes image artifacts more prominent, and a solution to this problem is desired. Depending on the training of the convolutional neural network, DLR may change the imaging features of lesions and obscure small lesions. Therefore, radiologists may need to adopt the habit of questioning whether any information has been lost on images that appear clean. ^©RSNA, 2023 Quiz questions for this article are available in the supplemental material.

Reconstruction for 7T high-resolution whole-brain diffusion MRI using two-stage N/2 ghost correction and L1-SPIRiT without single-band reference

PURPOSE

To combine a new two-stage N/2 ghost correction and an adapted L1-SPIRiT method for reconstruction of 7T highly accelerated whole-brain diffusion MRI (dMRI) using only autocalibration scans (ACS) without the need of additional single-band reference (SBref) scans.

METHODS

The proposed ghost correction consisted of a 3-line reference approach in stage 1 and the reference-free entropy method in stage 2. The adapted L1-SPIRiT method was formulated within the 3D k-space framework. Its efficacy was examined by acquiring two dMRI data sets at 1.05-mm isotropic resolutions with a total acceleration of 6 or 9 (i.e., 2-fold or 3-fold slice and 3-fold in-plane acceleration). Diffusion analysis was performed to derive DTI metrics and estimate fiber orientation distribution functions (fODFs). The results were compared with those of 3D k-space GRAPPA using only ACS, all in reference to 3D k-space GRAPPA using both ACS and SBref (serving as a reference).

RESULTS

The proposed ghost correction eliminated artifacts more robustly than conventional approaches. Our adapted L1-SPIRiT method outperformed 3D k-space GRAPPA when using only ACS, improving image quality to what was achievable with 3D k-space GRAPPA using both ACS and SBref scans. The improvement in image quality further resulted in an improvement in estimation performances for DTI and fODFs.

CONCLUSION

The combination of our new ghost correction and adapted L1-SPIRiT method can reliably reconstruct 7T highly accelerated whole-brain dMRI without the need of SBref scans, increasing acquisition efficiency and reducing motion sensitivity.

Reproducibility of non-contrast enhanced multi breath-hold ultrashort echo time functional lung MRI

PURPOSE

To evaluate the intraindividual reproducibility of functional lung imaging using non-contrast enhanced multi breath-hold 3D-UTE MRI.

METHODS

Ten healthy volunteers underwent non-contrast enhanced 3D-UTE MRI at three time points for same-day and different-day measurements employing a stack-of-spirals trajectory at 3 T. At each time point, inspiratory and expiratory breathing states were acquired for tidal and deep breathing, each within a single breath-hold. For functional image analysis, fractional ventilation (FV) was calculated pixelwise after image registration from the MR signal change. To decouple FV from breathing depth, the individual lung volume was used for volume adjustment (rFV). Reproducibility evaluation was performed in eight lung segments. Statistical analyses included two way mixed intraclass correlation (ICC), sign-test, Friedman-test and modified Bland-Altman analyses.

RESULTS

FV from tidal breathing showed an ICC of 0.81, a bias of 1.3% and an interval of confidence (CI) ranging from -67.1 to 69.6%. FV from deep breathing was higher reproducible with an ICC of 0.92 (bias, -0.2%; CI, -34.2 to 33.7%). Following volume adjustment, reproducibility of rFV for tidal breathing improved (ICC, 0,86; bias, 2.0%; CI, -34.3 to 38.3%), whereas it did not bear significant benefits for deep breathing (ICC, 0.89; bias, 2.8%; CI, -24.9 to 30.5%). Reproducibility was independent from the examination day.

CONCLUSION

Non-contrast-enhanced multi breath-hold 3D-UTE MRI allows for highly reproducible ventilation imaging.

AI in MRI: Computational Frameworks for a Faster, Optimized, and Automated Imaging Workflow

Over the last decade, artificial intelligence (AI) has made an enormous impact on a wide range of fields, including science, engineering, informatics, finance, and transportation [...].

JSENSE-Pro: Joint sensitivity estimation and image reconstruction in parallel imaging using pre-learned subspaces of coil sensitivity functions

PURPOSE

To improve calibrationless parallel imaging using pre-learned subspaces of coil sensitivity functions.

THEORY AND METHODS

A subspace-based joint sensitivity estimation and image reconstruction method was developed for improved parallel imaging with no calibration data. Specifically, we proposed to use a probabilistic subspace model to capture prior information of the coil sensitivity functions from previous scans acquired using the same receiver system. Both the subspace basis and coefficient distributions were learned from a small set of training data. The learned subspace model was then incorporated into the regularized reconstruction formalism that includes a sparsity prior. The nonlinear optimization problem was solved using alternating minimization algorithm. Public fastMRI brain dataset was retrospectively undersampled by different schemes for performance evaluation of the proposed method.

RESULTS

With no calibration data, the proposed method consistently produced the most accurate coil sensitivity estimation and highest quality image reconstructions at all acceleration factors tested in comparison with state-of-the-art methods including JSENSE, DeepSENSE, P-LORAKS, and Sparse BLIP. Our results are comparable to or even better than those from SparseSENSE, which used calibration data for sensitivity estimation. The work also demonstrated that the probabilistic subspace model learned from T₂ w data can be generalized to aiding the reconstruction of FLAIR data acquired from the same receiver system.

CONCLUSION

A subspace-based method named JSENSE-Pro has been proposed for accelerated parallel imaging without the acquisition of companion calibration data. The method is expected to further enhance the practical utility of parallel imaging, especially in applications where calibration data acquisition is not desirable or limited.

Rapid whole-brain quantitative MT imaging

PURPOSE

To provide a robust whole-brain quantitative magnetization transfer (MT) imaging method that is not limited by long acquisition times.

METHODS

Two variants of a spiral 2D interleaved multi-slice spoiled gradient echo (SPGR) sequence are used for rapid quantitative MT imaging of the brain at 3 T. A dual flip angle, steady-state prepared, double-contrast method is used for combined B₁ and-T₁ mapping in combination with a single-contrast MT-prepared acquisition over a range of different saturation flip angles (50 deg to 850 deg) and offset frequencies (1 kHz and 10 kHz). Five sets (containing minimum 6 to maximum 18 scans) with different MT-weightings were acquired. In addition, main magnetic field inhomogeneities (ΔB₀) were measured from two Cartesian low-resolution 2D SPGR scans with different echo times. Quantitative MT model parameters were derived from all sets using a two-pool continuous-wave model analysis, yielding the pool-size ratio, F, their exchange rate, k_f, and their transverse relaxation time, T_2r.

RESULTS

Whole-brain quantitative MT imaging was feasible for all sets with total acquisition times ranging from 7:15 min down to 3:15 min. For accurate modeling, B₁-correction was essential for all investigated sets, whereas ΔB₀-correction showed limited bias for the observed maximum off-resonances at 3 T.

CONCLUSION

The combination of rapid B₁-T₁ mapping and MT-weighted imaging using a 2D multi-slice spiral SPGR research sequence offers excellent prospects for rapid whole-brain quantitative MT imaging in the clinical setting.

Multi-weight respecification of scan-specific learning for parallel imaging

Parallel imaging is widely used in magnetic resonance imaging as an acceleration technology. Traditional linear reconstruction methods in parallel imaging often suffer from noise amplification. Recently, a non-linear robust artificial-neural-network for k-space interpolation (RAKI) exhibits superior noise resilience over other linear methods. However, RAKI performs poorly at high acceleration rates and needs a large number of autocalibration signals as the training samples. In order to tackle these issues, we propose a multi-weight method that implements multiple weighting matrices on the under-sampled data, named MW-RAKI. Enforcing multiple weighted matrices on the measurements can effectively reduce the influence of noise and increase the data constraints. Furthermore, we incorporate the strategy of multiple weighting matrixes into a residual version of RAKI, and form MW-rRAKI. Experimental comparisons with the alternative methods demonstrated noticeably better reconstruction performances, particularly at high acceleration rates. With only 12.5% of the k-space data is available, the PSNR of MW-RAKI and MW-rRAKI is improved by about 3 dB and 4 dB compared to RAKI and rRAKI, respectively.

Region-focused multi-view transformer-based generative adversarial network for cardiac cine MRI reconstruction

Cardiac cine magnetic resonance imaging (MRI) reconstruction is challenging due to spatial and temporal resolution trade-offs. Temporal correlation in cardiac cine MRI is informative and vital for understanding cardiac dynamic motion. Exploiting the temporal correlations in cine reconstruction is crucial to resolve aliasing artifacts and maintaining the cardiac motion patterns. However, existing methods have the following shortcomings: (1) they simultaneously compute pairwise correlations along spatial and temporal dimensions to establish dependencies, ignoring that learning spatial contextual information first will benefit the temporal modeling. (2) most studies neglect to focus on reconstructing the local cardiac regions, resulting in insufficient reconstruction accuracy due to a relatively large field of view. To address these problems, we propose a region-focused multi-view transformer-based generative adversarial network for cardiac cine MRI reconstruction. The proposed transformer divides consecutive cardiac frames into multiple views for cross-view feature extraction, establishing long-distance dependencies among features and effectively learning the spatio-temporal information. We further design a cross-view attention for spatio-temporal information fusion, ensuring the interaction of different spatio-temporal information in each view and capturing more temporal correlations of the cardiac motion. In addition, we introduce a cardiac region detection loss for improving the reconstruction quality of the cardiac region. Experimental results demonstrated that our method outperforms state-of-the-art methods. Especially with an acceleration factor as high as 10×, our method can reconstruct images with better accuracy and perceptual quality.

FFVN: An explicit feature fusion-based variational network for accelerated multi-coil MRI reconstruction

Magnetic Resonance Imaging (MRI) is a leading diagnostic imaging modality that supports high contrast of soft tissues with no invasiveness or radiation. Nonetheless, it suffers from long scan time owing to the inherent physics in its data acquisition process, hampering its development and applications. Traditional strategies such as Compressed Sensing (CS) and Parallel Imaging (PI) allow for MRI acceleration via sub-sampling strategy, and multiple coils, respectively. When Deep Learning (DL) joins in, both strategies get re-vitalized to achieve even faster reconstruction in various reconstruction methods, among which the variational network is a previously proposed method that combines the mathematical structure of variational models with DL for fast MRI reconstruction. However, in our study we observe that the information of MR features is either not efficiently or explicitly exploited in former works based on the variational network. Instead, we introduce a variational network with explicit feature fusion that combines the CS, PI, with DL for accelerated multi-coil MRI reconstruction. By explicitly leveraging the extra information via feature fusion following feature extraction, our proposed method achieves comparably satisfying performance to the state-of-the-art methods without too much computation overhead on a public multi-coil brain dataset under 5-fold and 10-fold acceleration.

A 3D stack-of-spirals approach for rapid hyperpolarized 129 Xe ventilation mapping in pediatric cystic fibrosis lung disease

PURPOSE

To demonstrate the feasibility of a rapid 3D stack-of-spirals (3D-SoS) imaging acquisition for hyperpolarized ¹²⁹ Xe ventilation mapping in healthy pediatric participants and pediatric cystic fibrosis (CF) participants, in comparison to conventional Cartesian multislice (2D) gradient-recalled echo (GRE) imaging.

METHODS

The 2D-GRE and 3D-SoS acquisitions were performed in 13 pediatric participants (5 healthy, 8 CF) during separate breath-holds. Images from both sequences were compared on the basis of ventilation defect percent (VDP) and other measures of image similarity. The nadir of transient oxygen saturation (SpO₂ ) decline due to xenon breath-holding was measured with pulse oximetry, and expressed as a percent change relative to baseline.

RESULTS

¹²⁹ Xe ventilation images were acquired in a breath-hold of 1.2-1.8 s with the 3D-SoS sequence, compared to 6.2-8.8 s for 2D-GRE. Mean ± SD VDP measures for 2D-GRE and 3D-SoS sequences were 5.02 ± 1.06% and 5.28 ± 1.08% in healthy participants, and 18.05 ± 8.26% and 18.75 ± 6.74% in CF participants, respectively. Across all participants, the intraclass correlation coefficient of VDP measures for both sequences was 0.98 (95% confidence interval: 0.94-0.99). The percent change in SpO₂ was reduced to -2.1 ± 2.7% from -5.2 ± 3.5% with the shorter 3D-SoS breath-hold.

CONCLUSION

Hyperpolarized ¹²⁹ Xe ventilation imaging with 3D-SoS yielded images approximately five times faster than conventional 2D-GRE, reducing SpO₂ desaturation and improving tolerability of the xenon administration. Analysis of VDP and other measures of image similarity demonstrate excellent agreement between images obtained with both sequences. 3D-SoS holds significant potential for reducing the acquisition time of hyperpolarized ¹²⁹ Xe MRI, and/or increasing spatial resolution while adhering to clinical breath-hold constraints.