Accelerating quantitative MR imaging with the incorporation of B1 compensation using deep learning

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.

B1 mapping using pre-learned subspaces for quantitative brain imaging

PURPOSE

To develop a machine learning-based method for estimation of both transmitter and receiver B1 fields desired for correction of the B1 inhomogeneity effects in quantitative brain imaging.

THEORY AND METHODS

A subspace model-based machine learning method was proposed for estimation of B1t and B1r fields. Probabilistic subspace models were used to capture scan-dependent variations in the B1 fields; the subspace basis and coefficient distributions were learned from pre-scanned training data. Estimation of the B1 fields for new experimental data was achieved by solving a linear optimization problem with prior distribution constraints. We evaluated the performance of the proposed method for B1 inhomogeneity correction in quantitative brain imaging scenarios, including T1 and proton density (PD) mapping from variable-flip-angle spoiled gradient-echo (SPGR) data as well as neurometabolic mapping from MRSI data, using phantom, healthy subject and brain tumor patient data.

RESULTS

In both phantom and healthy subject data, the proposed method produced high-quality B1 maps. B1 correction on SPGR data using the estimated B1 maps produced significantly improved T1 and PD maps. In brain tumor patients, the proposed method produced more accurate and robust B1 estimation and correction results than conventional methods. The B1 maps were also applied to MRSI data from tumor patients and produced improved neurometabolite maps, with better separation between pathological and normal tissues.

CONCLUSION

This work presents a novel method to estimate B1 variations using probabilistic subspace models and machine learning. The proposed method may make correction of B1 inhomogeneity effects more robust in practical applications.

Rapid estimation of 2D relative B 1 + -maps from localizers in the human heart at 7T using deep learning

PURPOSE

Subject-tailored parallel transmission pulses for ultra-high fields body applications are typically calculated based on subject-specific B 1 + $$ {\mathrm{B}}_1^{+} $$ -maps of all transmit channels, which require lengthy adjustment times. This study investigates the feasibility of using deep learning to estimate complex, channel-wise, relative 2D B 1 + $$ {\mathrm{B}}_1^{+} $$ -maps from a single gradient echo localizer to overcome long calibration times.

METHODS

126 channel-wise, complex, relative 2D B 1 + $$ {\mathrm{B}}_1^{+} $$ -maps of the human heart from 44 subjects were acquired at 7T using a Cartesian, cardiac gradient-echo sequence obtained under breath-hold to create a library for network training and cross-validation. The deep learning predicted maps were qualitatively compared to the ground truth. Phase-only B 1 + $$ {\mathrm{B}}_1^{+} $$ -shimming was subsequently performed on the estimated B 1 + $$ {\mathrm{B}}_1^{+} $$ -maps for a region of interest covering the heart. The proposed network was applied at 7T to 3 unseen test subjects.

RESULTS

The deep learning-based B 1 + $$ {\mathrm{B}}_1^{+} $$ -maps, derived in approximately 0.2 seconds, match the ground truth for the magnitude and phase. The static, phase-only pulse design performs best when maximizing the mean transmission efficiency. In-vivo application of the proposed network to unseen subjects demonstrates the feasibility of this approach: the network yields predicted B 1 + $$ {\mathrm{B}}_1^{+} $$ -maps comparable to the acquired ground truth and anatomical scans reflect the resulting B 1 + $$ {\mathrm{B}}_1^{+} $$ -pattern using the deep learning-based maps.

CONCLUSION

The feasibility of estimating 2D relative B 1 + $$ {\mathrm{B}}_1^{+} $$ -maps from initial localizer scans of the human heart at 7T using deep learning is successfully demonstrated. Because the technique requires only sub-seconds to derive channel-wise B 1 + $$ {\mathrm{B}}_1^{+} $$ -maps, it offers high potential for advancing clinical body imaging at ultra-high fields.

Deep Learning-Based Water-Fat Separation from Dual-Echo Chemical Shift-Encoded Imaging

Conventional water–fat separation approaches suffer long computational times and are prone to water/fat swaps. To solve these problems, we propose a deep learning-based dual-echo water–fat separation method. With IRB approval, raw data from 68 pediatric clinically indicated dual echo scans were analyzed, corresponding to 19382 contrast-enhanced images. A densely connected hierarchical convolutional network was constructed, in which dual-echo images and corresponding echo times were used as input and water/fat images obtained using the projected power method were regarded as references. Models were trained and tested using knee images with 8-fold cross validation and validated on out-of-distribution data from the ankle, foot, and arm. Using the proposed method, the average computational time for a volumetric dataset with ~400 slices was reduced from 10 min to under one minute. High fidelity was achieved (correlation coefficient of 0.9969, l1 error of 0.0381, SSIM of 0.9740, pSNR of 58.6876) and water/fat swaps were mitigated. I is of particular interest that metal artifacts were substantially reduced, even when the training set contained no images with metallic implants. Using the models trained with only contrast-enhanced images, water/fat images were predicted from non-contrast-enhanced images with high fidelity. The proposed water–fat separation method has been demonstrated to be fast, robust, and has the added capability to compensate for metal artifacts.

qMRI-BIDS: An extension to the brain imaging data structure for quantitative magnetic resonance imaging data

The Brain Imaging Data Structure (BIDS) established community consensus on the organization of data and metadata for several neuroimaging modalities. Traditionally, BIDS had a strong focus on functional magnetic resonance imaging (MRI) datasets and lacked guidance on how to store multimodal structural MRI datasets. Here, we present and describe the BIDS Extension Proposal 001 (BEP001), which adds a range of quantitative MRI (qMRI) applications to the BIDS. In general, the aim of qMRI is to characterize brain microstructure by quantifying the physical MR parameters of the tissue via computational, biophysical models. By proposing this new standard, we envision standardization of qMRI through multicenter dissemination of interoperable datasets. This way, BIDS can act as a catalyst of convergence between qMRI methods development and application-driven neuroimaging studies that can help develop quantitative biomarkers for neural tissue characterization. In conclusion, this BIDS extension offers a common ground for developers to exchange novel imaging data and tools, reducing the entrance barrier for qMRI in the field of neuroimaging.

Accelerating multi-echo chemical shift encoded water-fat MRI using model-guided deep learning

PURPOSE

To accelerate chemical shift encoded (CSE) water-fat imaging by applying a model-guided deep learning water-fat separation (MGDL-WF) framework to the undersampled k-space data.

METHODS

A model-guided deep learning water-fat separation framework is proposed for the acceleration using Cartesian/radial undersampling data. The proposed MGDL-WF combines the power of CSE water-fat imaging model and data-driven deep learning by jointly using a multi-peak fat model and a modified residual U-net network. The model is used to guide the image reconstruction, and the network is used to capture the artifacts induced by the undersampling. A data consistency layer is used in MGDL-WF to ensure the output images to be consistent with the k-space measurements. A Gauss-Newton iteration algorithm is adapted for the gradient updating of the networks.

RESULTS

Compared with the compressed sensing water-fat separation (CS-WF) algorithm/2-step procedure algorithm, the MGDL-WF increased peak signal-to-noise ratio (PSNR) by 5.31/5.23, 6.11/4.54, and 4.75 dB/1.88 dB with Cartesian sampling, and by 4.13/6.53, 2.90/4.68, and 1.68 dB/3.48 dB with radial sampling, at acceleration rates (R) of 4, 6, and 8, respectively. By using MGDL-WF, radial sampling increased the PSNR by 2.07 dB at R = 8, compared with Cartesian sampling.

CONCLUSIONS

The proposed MGDL-WF enables exploiting features of the water images and fat images from the undersampled multi-echo data, leading to improved performance in the accelerated CSE water-fat imaging. By using MGDL-WF, radial sampling can further improve the image quality with comparable scan time in comparison with Cartesian sampling.

Rigid motion-resolved B 1 + prediction using deep learning for real-time parallel-transmission pulse design

Purpose

Tailored parallel-transmit (pTx) pulses produce uniform excitation profiles at 7 T, but are sensitive to head motion. A potential solution is real-time pulse redesign. A deep learning framework is proposed to estimate pTx B₁⁺ distributions following within-slice motion, which can then be used for tailored pTx pulse redesign.

Methods

Using simulated data, conditional generative adversarial networks were trained to predict B₁⁺ distributions in the head following a displacement. Predictions were made for two virtual body models that were not included in training. Predicted maps were compared with groundtruth (simulated, following motion) B₁⁺ maps. Tailored pTx pulses were designed using B₁⁺ maps at the original position (simulated, no motion) and evaluated using simulated B₁⁺ maps at displaced position (ground-truth maps) to quantify motion-related excitation error. A second pulse was designed using predicted maps (also evaluated on ground-truth maps) to investigate improvement offered by the proposed method.

Results

Predicted B₁⁺ maps corresponded well with ground-truth maps. Error in predicted maps was lower than motion-related error in 99% and 67% of magnitude and phase evaluations, respectively. Worst-case flip-angle normalized RMS error due to motion (76% of target flip angle) was reduced by 59% when pulses were redesigned using predicted maps.

Conclusion

We propose a framework for predicting B₁⁺ maps online with deep neural networks. Predicted maps can then be used for real-time tailored pulse redesign, helping to overcome head motion–related error in pTx.

Synergistic multi-contrast cardiac magnetic resonance image reconstruction

Cardiac magnetic resonance imaging (CMR) is an important tool for the non-invasive diagnosis of a variety of cardiovascular diseases. Parametric mapping with multi-contrast CMR is able to quantify tissue alterations in myocardial disease and promises to improve patient care. However, magnetic resonance imaging is an inherently slow imaging modality, resulting in long acquisition times for parametric mapping which acquires a series of cardiac images with different contrasts for signal fitting or dictionary matching. Furthermore, extra efforts to deal with respiratory and cardiac motion by triggering and gating further increase the scan time. Several techniques have been developed to speed up CMR acquisitions, which usually acquire less data than that required by the Nyquist-Shannon sampling theorem, followed by regularized reconstruction to mitigate undersampling artefacts. Recent advances in CMR parametric mapping speed up CMR by synergistically exploiting spatial-temporal and contrast redundancies. In this article, we will review the recent developments in multi-contrast CMR image reconstruction for parametric mapping with special focus on low-rank and model-based reconstructions. Deep learning-based multi-contrast reconstruction has recently been proposed in other magnetic resonance applications. These developments will be covered to introduce the general methodology. Current technical limitations and potential future directions are discussed. This article is part of the theme issue 'Synergistic tomographic image reconstruction: part 1'.

Magnetic resonance parameter mapping using model-guided self-supervised deep learning

PURPOSE

To develop a model-guided self-supervised deep learning MRI reconstruction framework called reference-free latent map extraction (RELAX) for rapid quantitative MR parameter mapping.

METHODS

Two physical models are incorporated for network training in RELAX, including the inherent MR imaging model and a quantitative model that is used to fit parameters in quantitative MRI. By enforcing these physical model constraints, RELAX eliminates the need for full sampled reference data sets that are required in standard supervised learning. Meanwhile, RELAX also enables direct reconstruction of corresponding MR parameter maps from undersampled k-space. Generic sparsity constraints used in conventional iterative reconstruction, such as the total variation constraint, can be additionally included in the RELAX framework to improve reconstruction quality. The performance of RELAX was tested for accelerated T1 and T2 mapping in both simulated and actually acquired MRI data sets and was compared with supervised learning and conventional constrained reconstruction for suppressing noise and/or undersampling-induced artifacts.

RESULTS

In the simulated data sets, RELAX generated good T1 /T2 maps in the presence of noise and/or undersampling artifacts, comparable to artifact/noise-free ground truth. The inclusion of a spatial total variation constraint helps improve image quality. For the in vivo T1 /T2 mapping data sets, RELAX achieved superior reconstruction quality compared with conventional iterative reconstruction, and similar reconstruction performance to supervised deep learning reconstruction.

CONCLUSION

This work has demonstrated the initial feasibility of rapid quantitative MR parameter mapping based on self-supervised deep learning. The RELAX framework may also be further extended to other quantitative MRI applications by incorporating corresponding quantitative imaging models.

A practical method for post-acquisition reduction of bias in fast, whole-brain B1-maps

Large consistent differences have been observed between maps of the flip angle correction factor (commonly called "B₁-maps") produced with different fast methods in the human brain. We present an empirical procedure for first-order multiplicative bias correction that can be applied when more than one B₁-mapping method is available. We use a B₁-map measurement in a calibration phantom as a reference and the voxel-wise histogram mode between ratios of B₁-maps produced from different methods to calculate determine the bias as a multiplicative correcting scale factor. Institutional implementations of four common methods of B₁-mapping were assessed: Method of Slopes, FSE and EPI double angle methods (DAM), and Bloch-Siegert. In human subjects, the multiplicative bias used to correct for each of the four methods was: Method of Slopes = 1.005, FSE-DAM = 0.956, EPI-DAM = 1.080, and Bloch-Siegert = 1.128. Scaling to remove this bias between methods produces more consistent B₁-maps which enable more consistent values for any computations requiring flip angle correction. In addition, we present evidence that the corrected B₁ maps, using our calibration method, are also more accurate.

Deriving new soft tissue contrasts from conventional MR images using deep learning

Versatile soft tissue contrast in magnetic resonance imaging is a unique advantage of the imaging modality. However, the versatility is not fully exploited. In this study, we propose a deep learning-based strategy to derive more soft tissue contrasts from conventional MR images obtained in standard clinical MRI. Two types of experiments are performed. First, MR images corresponding to different pulse sequences are predicted from one or more images already acquired. As an example, we predict T_1ρ weighted knee image from T₂ weighted image and/or T₁ weighted image. Furthermore, we estimate images corresponding to alternative imaging parameter values. In a representative case, variable flip angle images are predicted from a single T₁ weighted image, whose accuracy is further validated in quantitative T₁ map subsequently derived. To accomplish these tasks, images are retrospectively collected from 56 subjects, and self-attention convolutional neural network models are trained using 1104 knee images from 46 subjects and tested using 240 images from 10 other subjects. High accuracy has been achieved in resultant qualitative images as well as quantitative T₁ maps. The proposed deep learning method can be broadly applied to obtain more versatile soft tissue contrasts without additional scans or used to normalize MR data that were inconsistently acquired for quantitative analysis.