snapMRF: GPU-accelerated magnetic resonance fingerprinting dictionary generation and matching using extended phase graphs

GPU-accelerated Bloch simulations and MR-STAT reconstructions using the Julia programming language

PURPOSE

MR-STAT is a relatively new multiparametric quantitative MRI technique in which quantitative paramater maps are obtained by solving a large-scale nonlinear optimization problem. Managing reconstruction times is one of the main challenges of MR-STAT. In this work we leverage GPU hardware to reduce MR-STAT reconstruction times. A highly optimized, GPU-compatible Bloch simulation toolbox is developed as part of this work that can be utilized for other quantitative MRI techniques as well.

METHODS

The Julia programming language was used to develop a flexible yet highly performant and GPU-compatible Bloch simulation toolbox called BlochSimulators.jl. The runtime performance of the toolbox is benchmarked against other Bloch simulation toolboxes. Furthermore, a (partially matrix-free) modification of a previously presented (matrix-free) MR-STAT reconstruction algorithm is proposed and implemented using the Julia language on GPU hardware. The proposed algorithm is combined with BlochSimulators.jl and the resulting MR-STAT reconstruction times on GPU hardware are compared to previously presented MR-STAT reconstruction times.

RESULTS

The BlochSimulators.jl package demonstrates superior runtime performance on both CPU and GPU hardware when compared to other existing Bloch simulation toolboxes. The GPU-accelerated partially matrix-free MR-STAT reconstruction algorithm, which relies on BlochSimulators.jl, allows for reconstructions of 68 seconds per two-dimensional (2D slice).

CONCLUSION

By combining the proposed Bloch simulation toolbox and the partially matrix-free reconstruction algorithm, 2D MR-STAT reconstructions can be performed in the order of one minute on a modern GPU card. The Bloch simulation toolbox can be utilized for other quantitative MRI techniques as well, for example for online dictionary generation for MR Fingerprinting.

SSL-QALAS: Self-Supervised Learning for rapid multiparameter estimation in quantitative MRI using 3D-QALAS

PURPOSE

To develop and evaluate a method for rapid estimation of multiparametric T1 , T2 , proton density, and inversion efficiency maps from 3D-quantification using an interleaved Look-Locker acquisition sequence with T2 preparation pulse (3D-QALAS) measurements using self-supervised learning (SSL) without the need for an external dictionary.

METHODS

An SSL-based QALAS mapping method (SSL-QALAS) was developed for rapid and dictionary-free estimation of multiparametric maps from 3D-QALAS measurements. The accuracy of the reconstructed quantitative maps using dictionary matching and SSL-QALAS was evaluated by comparing the estimated T1 and T2 values with those obtained from the reference methods on an International Society for Magnetic Resonance in Medicine/National Institute of Standards and Technology phantom. The SSL-QALAS and the dictionary-matching methods were also compared in vivo, and generalizability was evaluated by comparing the scan-specific, pre-trained, and transfer learning models.

RESULTS

Phantom experiments showed that both the dictionary-matching and SSL-QALAS methods produced T1 and T2 estimates that had a strong linear agreement with the reference values in the International Society for Magnetic Resonance in Medicine/National Institute of Standards and Technology phantom. Further, SSL-QALAS showed similar performance with dictionary matching in reconstructing the T1 , T2 , proton density, and inversion efficiency maps on in vivo data. Rapid reconstruction of multiparametric maps was enabled by inferring the data using a pre-trained SSL-QALAS model within 10 s. Fast scan-specific tuning was also demonstrated by fine-tuning the pre-trained model with the target subject's data within 15 min.

CONCLUSION

The proposed SSL-QALAS method enabled rapid reconstruction of multiparametric maps from 3D-QALAS measurements without an external dictionary or labeled ground-truth training data.

Feasibility of joint mapping of triglyceride saturation and water longitudinal relaxation in a single breath hold applied to high fat-fraction adipose depots in the periclavicular anatomy

INTRODUCTION

Simultaneous mapping of triglyceride (TAG) saturation and tissue water relaxation may improve the characterization of the structure and function of anatomies with significant adipose tissue. While several groups have demonstrated in vivo TAG saturation imaging using MRI, joint mapping of relaxation and TAG saturation is understudied. Such mappings may avoid bias from physiological motion, if they can be done within a single breath-hold, and also account for static and applied magnetic field heterogeneity.

METHODS

We propose a transient-state/MR fingerprinting single breath-hold sequence at 3 T, a low-rank reconstruction, and a parameter estimation pipeline that jointly estimates the number of double bonds (NDB), number of methylene interrupted double bonds (NMIDB), and tissue water T1, while accounting for non-ideal radiofrequency transmit scaling and off-resonance effects. We test the proposed method in simulations, in phantom against MR spectroscopy (MRS), and in vivo regions in and around high fat fraction (FF) periclavicular adipose tissue. Partial volume and multi-peak transverse relaxation effects are explored.

RESULTS

The simulation results demonstrate accurate NDB, NMIDB, and water T1 estimates across a range of NDB, NMIDB, and T1 values. In phantoms, the proposed method's estimates of NDB and NMIDB correlate with those from MR spectroscopy (Pearson correlation ≥0.98), while the water T1 estimates are concordant with a standard phantom. The NDB and NMIDB are sensitive to partial volumes of water, showing increasing bias at FF < 40%. This bias is found to be due to noise and transverse relaxation effects. The in vivo periclavicular adipose tissue has high FF (>90%). The adipose tissue NDB and NMIDB, and muscle T1 estimates are comparable to those reported in the literature.

CONCLUSION

Robust estimation of NDB, NMIDB at high FF and water T1 across a broad range of FFs are feasible using the proposed methods. Further reduction of noise and model bias are needed to employ the proposed technique in low FF anatomies and pathologies.

Artificial intelligence in cardiac magnetic resonance fingerprinting

Magnetic resonance fingerprinting (MRF) is a fast MRI-based technique that allows for multiparametric quantitative characterization of the tissues of interest in a single acquisition. In particular, it has gained attention in the field of cardiac imaging due to its ability to provide simultaneous and co-registered myocardial T₁ and T₂ mapping in a single breath-held cardiac MRF scan, in addition to other parameters. Initial results in small healthy subject groups and clinical studies have demonstrated the feasibility and potential of MRF imaging. Ongoing research is being conducted to improve the accuracy, efficiency, and robustness of cardiac MRF. However, these improvements usually increase the complexity of image reconstruction and dictionary generation and introduce the need for sequence optimization. Each of these steps increase the computational demand and processing time of MRF. The latest advances in artificial intelligence (AI), including progress in deep learning and the development of neural networks for MRI, now present an opportunity to efficiently address these issues. Artificial intelligence can be used to optimize candidate sequences and reduce the memory demand and computational time required for reconstruction and post-processing. Recently, proposed machine learning-based approaches have been shown to reduce dictionary generation and reconstruction times by several orders of magnitude. Such applications of AI should help to remove these bottlenecks and speed up cardiac MRF, improving its practical utility and allowing for its potential inclusion in clinical routine. This review aims to summarize the latest developments in artificial intelligence applied to cardiac MRF. Particularly, we focus on the application of machine learning at different steps of the MRF process, such as sequence optimization, dictionary generation and image reconstruction.

MR fingerprinting of the prostate

Multiparametric magnetic resonance imaging (mpMRI) has been adopted as the key tool for detection, localization, characterization, and risk stratification of patients suspected to have prostate cancer. Despite advantages over systematic biopsy, the interpretation of prostate mpMRI has limitations including a steep learning curve, leading to considerable interobserver variation. There is growing interest in clinical translation of quantitative imaging techniques for more objective lesion assessment. However, traditional mapping techniques are slow, precluding their use in the clinic. Magnetic resonance fingerprinting (MRF) is an efficient approach for quantitative maps of multiple tissue properties simultaneously. The T₁ and T₂ values obtained with MRF have been validated with phantom studies as well as in normal volunteers and patients. Studies have shown that MRF-derived T₁ and T₂ along with ADC values are all significant independent predictors in the differentiation between normal prostate tissue and prostate cancer, and hold promise in differentiating low and intermediate/high-grade cancers. This review seeks to introduce the basics of the prostate MRF technique, discuss the potential applications of prostate MRF for the characterization of prostate cancer, and describes ongoing areas of research.

Constrained Ellipse Fitting for Efficient Parameter Mapping With Phase-Cycled bSSFP MRI

Balanced steady-state free precession (bSSFP) imaging enables high scan efficiency in MRI, but differs from conventional sequences in terms of elevated sensitivity to main field inhomogeneity and nonstandard [Formula: see text]-weighted tissue contrast. To address these limitations, multiple bSSFP images of the same anatomy are commonly acquired with a set of different RF phase-cycling increments. Joint processing of phase-cycled acquisitions serves to mitigate sensitivity to field inhomogeneity. Recently phase-cycled bSSFP acquisitions were also leveraged to estimate relaxation parameters based on explicit signal models. While effective, these model-based methods often involve a large number of acquisitions (N ≈ 10-16), degrading scan efficiency. Here, we propose a new constrained ellipse fitting method (CELF) for parameter estimation with improved efficiency and accuracy in phase-cycled bSSFP MRI. CELF is based on the elliptical signal model framework for complex bSSFP signals; and it introduces geometrical constraints on ellipse properties to improve estimation efficiency, and dictionary-based identification to improve estimation accuracy. CELF generates maps of [Formula: see text], [Formula: see text], off-resonance and on-resonant bSSFP signal by employing a separate [Formula: see text] map to mitigate sensitivity to flip angle variations. Our results indicate that CELF can produce accurate off-resonance and banding-free bSSFP maps with as few as N = 4 acquisitions, while estimation accuracy for relaxation parameters is notably limited by biases from microstructural sensitivity of bSSFP imaging.

Accelerated reconstruction of dictionary-based T2 relaxation maps based on dictionary compression and gradient descent search algorithms

Background Quantitative T₂-relaxation-based contrast maps have shown to be highly beneficial for clinical diagnosis and follow-up. The generation of quantitative maps, however, is impaired by long acquisition times, and time-consuming post-processing schemes. The EMC platform is a dictionary-based technique, which involves simulating theoretical signal curves for different physical and experimental values, followed by matching the experimentally acquired signals to the set simulated ones. Purpose Although the EMC technique has shown to produce accurate T₂ maps, it involves computationally intensive post-processing procedures. In this work we present an approach for accelerating the reconstruction of T₂ relaxation maps. Methods This work presents two alternative post-processing approaches for accelerating the reconstruction of EMC-based T₂ relaxation maps. These are (a) Dictionary compression using principal component analysis (PCA) and (b) gradient-descent search algorithm. Additional acceleration was achieved by finding the optimal MATLAB C++ compiler. The utility of the two suggested approaches was examined by calculating the relative error, produced by each technique. Results Gradient descent method was in perfect agreement with the ground truth exhaustive search matching process. PCA based acceleration produced root mean square error (RMSE) of up to 4% compared to exhaustive matching process. Overall acceleration of x16 was achieved using gradient descent in addition to x7 acceleration by choosing the optimal MATLAB C++ compiler. Conclusions Postprocessing of EMC-based T₂ relaxation maps can be accelerated without impairing the accuracy of the ensuing T₂ values.

A multi-inversion multi-echo spin and gradient echo echo planar imaging sequence with low image distortion for rapid quantitative parameter mapping and synthetic image contrasts

PURPOSE

Brain imaging exams typically take 10-20 min and involve multiple sequential acquisitions. A low-distortion whole-brain echo planar imaging (EPI)-based approach was developed to efficiently encode multiple contrasts in one acquisition, allowing for calculation of quantitative parameter maps and synthetic contrast-weighted images.

METHODS

Inversion prepared spin- and gradient-echo EPI was developed with slice-order shuffling across measurements for efficient acquisition with T1 , T2 , and T 2 ∗ weighting. A dictionary-matching approach was used to fit the images to quantitative parameter maps, which in turn were used to create synthetic weighted images with typical clinical contrasts. Dynamic slice-optimized multi-coil shimming with a B0 shim array was used to reduce B0 inhomogeneity and, therefore, image distortion by >50%. Multi-shot EPI was also implemented to minimize distortion and blurring while enabling high in-plane resolution. A low-rank reconstruction approach was used to mitigate errors from shot-to-shot phase variation.

RESULTS

The slice-optimized shimming approach was combined with in-plane parallel-imaging acceleration of 4× to enable single-shot EPI with more than eight-fold distortion reduction. The proposed sequence efficiently obtained 40 contrasts across the whole-brain in just over 1 min at 1.2 × 1.2 × 3 mm resolution. The multi-shot variant of the sequence achieved higher in-plane resolution of 1 × 1 × 4 mm with good image quality in 4 min. Derived quantitative maps showed comparable values to conventional mapping methods.

CONCLUSION

The approach allows fast whole-brain imaging with quantitative parameter maps and synthetic weighted contrasts. The slice-optimized multi-coil shimming and multi-shot reconstruction approaches result in minimal EPI distortion, giving the sequence the potential to be used in rapid screening applications.

Fast and accurate modeling of transient-state, gradient-spoiled sequences by recurrent neural networks

Fast and accurate modeling of MR signal responses are typically required for various quantitative MRI applications, such as MR fingerprinting. This work uses a new extended phase graph (EPG)-Bloch model for accurate simulation of transient-state, gradient-spoiled MR sequences, and proposes a recurrent neural network (RNN) as a fast surrogate of the EPG-Bloch model for computing large-scale MR signals and derivatives. The computational efficiency of the RNN model is demonstrated by comparisons with other existing models, showing one to three orders of acceleration compared with the latest GPU-accelerated, open-source EPG package. By using numerical and in vivo brain data, two used cases, namely, MRF dictionary generation and optimal experimental design, are also provided. Results show that the RNN surrogate model can be efficiently used for computing large-scale dictionaries of transient-state signals and derivatives within tens of seconds, resulting in several orders of magnitude acceleration with respect to state-of-the-art implementations. The practical application of transient-state quantitative techniques can therefore be substantially facilitated.

Cardiac magnetic resonance fingerprinting: Trends in technical development and potential clinical applications

Quantitative cardiac magnetic resonance has emerged in recent years as an approach for evaluating a range of cardiovascular conditions, with T1 and T2 mapping at the forefront of these developments. Cardiac Magnetic Resonance Fingerprinting (cMRF) provides a rapid and robust framework for simultaneous quantification of myocardial T1 and T2 in addition to other tissue properties. Since the advent of cMRF, a number of technical developments and clinical validation studies have been reported. This review provides an overview of cMRF, recent technical developments, healthy subject and patient studies, anticipated technical improvements, and potential clinical applications. Recent technical developments include slice profile and pulse efficiency corrections, improvements in image reconstruction, simultaneous multislice imaging, 3D whole-ventricle imaging, motion-resolved imaging, fat-water separation, and machine learning for rapid dictionary generation. Future technical developments in cMRF, such as B0 and B1 field mapping, acceleration of acquisition and reconstruction, imaging of patients with implanted devices, and quantification of additional tissue properties are also described. Potential clinical applications include characterization of infiltrative, inflammatory, and ischemic cardiomyopathies, tissue characterization in the left atrium and right ventricle, post-cardiac transplantation assessment, reduction of contrast material, pre-procedural planning for electrophysiology interventions, and imaging of patients with implanted devices.

Model-informed machine learning for multi-component T2 relaxometry

Recovering the T₂ distribution from multi-echo T₂ magnetic resonance (MR) signals is challenging but has high potential as it provides biomarkers characterizing the tissue micro-structure, such as the myelin water fraction (MWF). In this work, we propose to combine machine learning and aspects of parametric (fitting from the MRI signal using biophysical models) and non-parametric (model-free fitting of the T₂ distribution from the signal) approaches to T₂ relaxometry in brain tissue by using a multi-layer perceptron (MLP) for the distribution reconstruction. For training our network, we construct an extensive synthetic dataset derived from biophysical models in order to constrain the outputs with a priori knowledge of in vivo distributions. The proposed approach, called Model-Informed Machine Learning (MIML), takes as input the MR signal and directly outputs the associated T₂ distribution. We evaluate MIML in comparison to a Gaussian Mixture Fitting (parametric) and Regularized Non-Negative Least Squares algorithms (non-parametric) on synthetic data, an ex vivo scan, and high-resolution scans of healthy subjects and a subject with Multiple Sclerosis. In synthetic data, MIML provides more accurate and noise-robust distributions. In real data, MWF maps derived from MIML exhibit the greatest conformity to anatomical scans, have the highest correlation to a histological map of myelin volume, and the best unambiguous lesion visualization and localization, with superior contrast between lesions and normal appearing tissue. In whole-brain analysis, MIML is 22 to 4980 times faster than the non-parametric and parametric methods, respectively.

Quantitative Skeletal Muscle Imaging Using 3D MR Fingerprinting With Water and Fat Separation

BACKGROUND

Quantitative muscle MRI is a robust tool to monitor intramuscular fatty replacement and disease activity in patients with neuromuscular disorders (NMDs).

PURPOSE

To implement a 3D sequence for quantifying simultaneously fat fraction (FF) and water T₁ (T_1,H2O ) in the skeletal muscle, evaluate regular undersampling in the partition-encoding direction, and compare it to a recently proposed 2D MR fingerprinting sequence with water and fat separation (MRF T₁ -FF).

STUDY TYPE

Prospective.

PHANTOM/SUBJECTS

Seventeen-vial phantom at different FF and T_1,H2O , 11 healthy volunteers, and 6 subjects with different NMDs.

FIELD STRENGTH/SEQUENCE

3T/3D MRF T₁ -FF, 2D MRF T₁ -FF, STEAM MRS ASSESSMENT: FF and T_1,H2O measured with the 2D and 3D sequences were compared in the phantom and in vivo at different undersampling factors (US). Data were acquired in healthy subjects before and after plantar dorsiflexions and at rest in patients.

STATISTICAL TESTS

Linear correlations, Bland-Altman analysis, two-way repeated measures analysis of variance (ANOVA), Student's t-test.

RESULTS

Up to a US factor of 3, the undersampled acquisitions were in good agreement with the fully sampled sequence (R²  ≥ 0.98, T_1,H2O bias ≤10 msec, FF bias ≤4 × 10^-4 ) both in phantom and in vivo. The 2D and 3D MRF T₁ -FF sequences provided comparable T_1,H2O and FF values (R²  ≥ 0.95, absolute T_1,H2O bias ≤35 msec, and absolute FF bias ≤0.003). The plantar dorsiflexion induced a significant increase of T_1,H2O in the tibialis anterior and extensor digitorum (relative increase of +10.8 ± 1.7% and + 7.7 ± 1.4%, respectively, P < 0.05), that was accompanied by a significant reduction of FF in the tibialis anterior (relative decrease of -16.3 ± 4.0%, P < 0.05). Some subjects with NMDs presented increased and heterogeneous T_1,H2O and FF values throughout the leg.

DATA CONCLUSION

Quantitative 3D T_1,H2O and FF maps covering the entire leg were obtained within acquisition times compatible with clinical research (4 minutes 20 seconds) and a 1 × 1 × 5 mm³ spatial resolution.

LEVEL OF EVIDENCE

2 TECHNICAL EFFICACY: Stage 2.

Slice-selective extended phase graphs in gradient-crushed, transient-state free precession sequences: An application to MR fingerprinting

PURPOSE

Slice-selective, gradient-crushed, transient-state sequences such as those used in MR fingerprinting (MRF) relaxometry are sensitive to slice profile effects. Whereas balanced steady-state free precession MRF profile effects have been studied, less attention has been given to gradient-crushed MRF forms. Extensions of the extended phase graph (EPG) formalism, called slice-selective EPG (ssEPG), are proposed that model slice profile effects.

THEORY AND METHODS

The hard-pulse approximation to slice-selective excitation in the spatial domain is reformulated in k-space. Excitation is modeled by standard EPG shift and transition operators. This ssEPG modeling is validated against Bloch simulations and phantom slice profile measurements. ssEPG relaxometry accuracy and variability are compared with other EPG methods in phantoms and human leg in vivo. The role of ∆B₀ interactions with slice profile and gradient crushers is investigated.

RESULTS

Simulations and slice profile measurements show that ssEPG can model highly dynamic slice profile effects of gradient-crushed sequences. The MRF ssEPG T₂ estimates over 0 < T₂ < 100 ms improve accuracy by > 10 ms at some values relative to other modeling approaches. Small deviations in B₀ can produce substantial bias in T₂ estimations from a range of MRF sequence types, and these effects can be modeled and understood by ssEPG.

CONCLUSION

Transient-state, gradient-crushed sequences such as those used in MRF are sensitive to slice profile effects, and these effects depend on RF pulse choice, gradient crusher strength, and ∆B₀ . It was found ssEPG was the most accurate EPG-based means to model these effects.