High resolution myelin water imaging incorporating local tissue susceptibility analysis

A predictor-corrector phase unwrapping algorithm for temporally undersampled gradient-echo MRI

PURPOSE

To develop a method for unwrapping temporally undersampled and nonlinear gradient recalled echo (GRE) phase.

THEORY AND METHODS

Temporal unwrapping is performed as a sequential one step prediction of the echo phase, followed by a correction to the nearest integer wrap-count. A spatio-temporal extension of the 1D predictor corrector unwrapping (PCU) algorithm improves the prediction accuracy, and thereby maintains spatial continuity. The proposed method is evaluated using numerical phantom, physical phantom, and in vivo brain data at both 3 T and 9.4 T. The unwrapping performance is compared with the state-of-the-art temporal and spatial unwrapping algorithms, and the spatio-temporal iterative virtual-echo based Nyquist sampled (iVENyS) algorithm.

RESULTS

Simulation results showed significant reduction in unwrapping errors at higher echoes compared with the state-of-the-art algorithms. Similar to the iVENyS algorithm, the PCU algorithm was able to generate spatially smooth phase images for in vivo data acquired at 3 T and 9.4 T, bypassing the use of additional spatial unwrapping step. A key advantage over iVENyS algorithm is the superior performance of PCU algorithm at higher echoes.

CONCLUSION

PCU algorithm serves as a robust phase unwrapping method for temporally undersampled and nonlinear GRE phase, particularly in the presence of high field gradients.

Optimized three-dimensional ultrashort echo time: Magnetic resonance fingerprinting for myelin tissue fraction mapping

Quantitative assessment of brain myelination has gained attention for both research and diagnosis of neurological diseases. However, conventional pulse sequences cannot directly acquire the myelin-proton signals due to its extremely short T2 and T2* values. To obtain the myelin-proton signals, dedicated short T2 acquisition techniques, such as ultrashort echo time (UTE) imaging, have been introduced. However, it remains challenging to isolate the myelin-proton signals from tissues with longer T2. In this article, we extended our previous two-dimensional ultrashort echo time magnetic resonance fingerprinting (UTE-MRF) with dual-echo acquisition to three dimensional (3D). Given a relatively low proton density (PD) of myelin-proton, we utilized Cramér-Rao Lower Bound to encode myelin-proton with the maximal SNR efficiency for optimizing the MR fingerprinting design, in order to improve the sensitivity of the sequence to myelin-proton. In addition, with a second echo of approximately 3 ms, myelin-water component can be also captured. A myelin-tissue (myelin-proton and myelin-water) fraction mapping can be thus calculated. The optimized 3D UTE-MRF with dual-echo acquisition is tested in simulations, physical phantom and in vivo studies of both healthy subjects and multiple sclerosis patients. The results suggest that the rapidly decayed myelin-proton and myelin-water signal can be depicted with UTE signals of our method at clinically relevant resolution (1.8 mm isotropic) in 15 min. With its good sensitivity to myelin loss in multiple sclerosis patients demonstrated, our method for the whole brain myelin-tissue fraction mapping in clinical friendly scan time has the potential for routine clinical imaging.

NeXtQSM-A complete deep learning pipeline for data-consistent Quantitative Susceptibility Mapping trained with hybrid data

Deep learning based Quantitative Susceptibility Mapping (QSM) has shown great potential in recent years, obtaining similar results to established non-learning approaches. Many current deep learning approaches are not data consistent, require in vivo training data or solve the QSM problem in consecutive steps resulting in the propagation of errors. Here we aim to overcome these limitations and developed a framework to solve the QSM processing steps jointly. We developed a new hybrid training data generation method that enables the end-to-end training for solving background field correction and dipole inversion in a data-consistent fashion using a variational network that combines the QSM model term and a learned regularizer. We demonstrate that NeXtQSM overcomes the limitations of previous deep learning methods. NeXtQSM offers a new deep learning based pipeline for computing quantitative susceptibility maps that integrates each processing step into the training and provides results that are robust and fast.

Improved magnetic resonance myelin water imaging using multi-channel denoising convolutional neural networks (MCDnCNN)

BACKGROUND

Myelin water imaging (MWI) is powerful and important for studying and diagnosing neurological and psychiatric diseases. In particular, myelin water fraction (MWF) is derived from MWI data for quantifying myelination. However, MWF estimation is typically sensitive to noise. Improving the accuracy of MWF estimation based on WMI data acquired using a magnetic resonance (MR) multiple gradient recalled echo (mGRE) imaging sequence is desired.

METHODS

The proposed method employs a recently introduced the multi-channel denoising convolutional neural networks (MCDnCNN). Five different MCDnCNN models, denoted as Delevel1, Delevel2, Delevel3, Delevel4 and DelevelMix corresponding to five noise levels (Level1, Level2, Level3, Level4 and LevelMix), were trained using the data of the first echo of the mGRE brain images acquired from 15 healthy human subjects. Using simulated noisy data that employed a hollow cylinder model, we first evaluated the improvement in estimating MWF based on data denoised by the five different MCDnCNNs, by comparing the MWF maps calculated from the denoised data with ground truth. Next, we again evaluated the improvement using real-world in vivo datasets of 11 human participants acquired using the mGRE sequence. The datasets were first denoised by five different MCDnCNNs (Delevel1, 2, 3, 4 and DelevelMix), and subsequently their MWF maps were calculated and compared with the MWF maps directly calculated from the raw mGRE images without being denoised.

RESULTS

Experiments using the simulation data denoised by the appropriate MCDnCNN models showed that the standard deviation (SD) of the absolute error (AE) of the derived MWF results was significantly reduced (maximal reduction =15.5%, Level3 simulated noisy data, orientation angle =0, all the five MCDnCNN models). In the test using in vivo data, estimating MWF based on data particularly denoised by the appropriate MCDnCNN models was found to be the best, compared to otherwise not using the appropriate models. The results demonstrated that the appropriate MCDnCNN models may permit high-quality MWF mapping, i.e., substantial reduction of random variation in estimating MWF-maps while preserving accuracy and structural details.

CONCLUSIONS

Appropriate MCDnCNN models as proposed may improve both the accuracy and precision in estimating MWF maps, thereby making it a more clinically feasible alternative.

Comparison of non-parametric T2 relaxometry methods for myelin water quantification

Multi-component T₂ relaxometry allows probing tissue microstructure by assessing compartment-specific T₂ relaxation times and water fractions, including the myelin water fraction. Non-negative least squares (NNLS) with zero-order Tikhonov regularization is the conventional method for estimating smooth T₂ distributions. Despite the improved estimation provided by this method compared to non-regularized NNLS, the solution is still sensitive to the underlying noise and the regularization weight. This is especially relevant for clinically achievable signal-to-noise ratios. In the literature of inverse problems, various well-established approaches to promote smooth solutions, including first-order and second-order Tikhonov regularization, and different criteria for estimating the regularization weight have been proposed, such as L-curve, Generalized Cross-Validation, and Chi-square residual fitting. However, quantitative comparisons between the available reconstruction methods for computing the T₂ distribution, and between different approaches for selecting the optimal regularization weight, are lacking. In this study, we implemented and evaluated ten reconstruction algorithms, resulting from the individual combinations of three penalty terms with three criteria to estimate the regularization weight, plus non-regularized NNLS. Their performance was evaluated both in simulated data and real brain MRI data acquired from healthy volunteers through a scan-rescan repeatability analysis. Our findings demonstrate the need for regularization. As a result of this work, we provide a list of recommendations for selecting the optimal reconstruction algorithms based on the acquired data. Moreover, the implemented methods were packaged in a freely distributed toolbox to promote reproducible research, and to facilitate further research and the use of this promising quantitative technique in clinical practice.

Evaluation of the diffusion MRI white matter tract integrity model using myelin histology and Monte-Carlo simulations

Quantitative evaluation of brain myelination has drawn considerable attention. Conventional diffusion-based magnetic resonance imaging models, including diffusion tensor imaging and diffusion kurtosis imaging (DKI),¹ have been used to infer the microstructure and its changes in neurological diseases. White matter tract integrity (WMTI) was proposed as a biophysical model to relate the DKI-derived metrics to the underlying microstructure. Although the model has been validated on ex vivo animal brains, it was not well evaluated with ex vivo human brains. In this study, histological samples (namely corpus callosum) from postmortem human brains have been investigated based on WMTI analyses on a clinical 3T scanner and comparisons with gold standard myelin staining in proteolipid protein and Luxol fast blue. In addition, Monte Carlo simulations were conducted to link changes from ex vivo to in vivo conditions based on the microscale parameters of water diffusivity and permeability. The results show that WMTI metrics, including axonal water fraction AWF, radial extra-axonal diffusivity D_e^⊥, and intra-axonal diffusivity D_awere needed to characterize myelin content alterations. Thus, WMTI model metrics are shown to be promising candidates as sensitive biomarkers of demyelination.

Artificial neural network for multi-echo gradient echo-based myelin water fraction estimation

PURPOSE

To demonstrate robust myelin water fraction (MWF) mapping using an artificial neural network (ANN) with multi-echo gradient-echo (GRE) signal.

METHODS

Multi-echo gradient-echo signals simulated with a three-pool exponential model were used to generate the training data set for the ANN, which was designed to yield the MWF. We investigated the performance of our proposed ANN for various conditions using both numerical simulations and in vivo data. Simulations were conducted with various SNRs to investigate the performance of the ANN. In vivo data with high spatial resolutions were applied in the analyses, and results were compared with MWFs derived by the nonlinear least-squares algorithm using a complex three-pool exponential model.

RESULTS

The network results for the simulations show high accuracies against noise compared with nonlinear least-squares MWFs: RMS-error value of 5.46 for the nonlinear least-squares MWF and 3.56 for the ANN MWF at an SNR of 150 (relative gain = 34.80%). These effects were also found in the in vivo data, with reduced SDs in the region-of-interest analyses. These effects of the ANN demonstrate the feasibility of acquiring high-resolution myelin water images.

CONCLUSION

The simulation results and in vivo data suggest that the ANN facilitates more robust MWF mapping in multi-echo gradient-echo sequences compared with the conventional nonlinear least-squares method.

The association of myelination in the internal capsule with iron deposition in the basal ganglia in macaques: a magnetic resonance imaging study

Background

Iron plays a vital role in myelin synthesis and maintenance. A tight association between iron concentration and myelin content has been demonstrated in local brain regions; however, whether such a relationship exists between distant brain regions that are anatomically connected is largely unknown.

Methods

We conducted an in vivo measurement of iron and myelin content in the brains of 8 young (mean age: 7.7 years) and 8 old (mean age: 24.7 years) macaques by integrating two MRI-based techniques: quantitative susceptibility mapping (QSM) and myelin water fraction (MWF) imaging. We examined the relationship between iron deposition in components of the basal ganglia (BG), and the myelin content of the BG-connecting fiber tract internal capsule (IC) and four more white matter (WM) structures, including the optic tract, and the genu, body, and splenium of the corpus callosum, which are anatomically separate from the BG.

Results

Spearman's correlation analysis revealed a moderate to high (r=0.528-0.808, P<0.05) positive correlation between the magnetic susceptibility of the BG and the MWF of anatomically connected IC structures during myelin production and maintenance, but little significant correlation was found between the susceptibility of the BG and the MWF of WM structures not anatomically connected to the BG.

Conclusions

These results advance the understanding of the relationship between iron and myelin, and suggest that future studies should consider the impact that iron concentration in the BG has on the myelination of WM structures that are anatomically connected to the BG.

Field strength influences on gradient recalled echo MRI signal compartment frequency shifts

Different trends of echo time dependent gradient recalled echo MRI signals in different brain regions have been attributed to signal compartments in image voxels. It remains unclear how variations in gradient recalled echo MRI signals change as a function of MRI field strength, and how data processing may impact signal compartment parameters. We used two popular quantitative susceptibility mapping methods of processing raw phase images (Laplacian and path-based unwrapping with V-SHARP) and expressed values in the form of induced frequency shifts (in Hz) in six specific brain regions at 3T and 7T. We found the frequency shift curves to vary with echo time, and a good overlap between 3T and 7T mean frequency shift curves was present. However, the amount of variation across participants was greater at 3T, and we were able to obtain better compartment model fits of the signal at 7T. We also found the temporal trends in the signal and compartment frequency shifts to change with the method used to process images. The inter-participant averaged trends were consistent between 3T and 7T for each quantitative susceptibility pipeline. However, signal compartment frequency shifts generated using different pipelines may not be comparable.

Multi-Exponential Relaxometry Using l1 -Regularized Iterative NNLS (MERLIN) With Application to Myelin Water Fraction Imaging

A new parameter estimation algorithm, MERLIN, is presented for accurate and robust multi-exponential relaxometry using magnetic resonance imaging, a tool that can provide valuable insight into the tissue microstructure of the brain. Multi-exponential relaxometry is used to analyze the myelin water fraction and can help to detect related diseases. However, the underlying problem is ill-conditioned, and as such, is extremely sensitive to noise and measurement imperfections, which can lead to less precise and more biased parameter estimates. MERLIN is a fully automated, multi-voxel approach that incorporates state-of-the-art l₁ -regularization to enforce sparsity and spatial consistency of the estimated distributions. The proposed method is validated in simulations and in vivo experiments, using a multi-echo gradient-echo (MEGE) sequence at 3 T. MERLIN is compared to the conventional single-voxel l₂ -regularized NNLS (rNNLS) and a multi-voxel extension with spatial priors (rNNLS + SP), where it consistently showed lower root mean squared errors of up to 70 percent for all parameters of interest in these simulations.

DeepQSM - using deep learning to solve the dipole inversion for quantitative susceptibility mapping

Quantitative susceptibility mapping (QSM) is based on magnetic resonance imaging (MRI) phase measurements and has gained broad interest because it yields relevant information on biological tissue properties, predominantly myelin, iron and calcium in vivo. Thereby, QSM can also reveal pathological changes of these key components in widespread diseases such as Parkinson's disease, Multiple Sclerosis, or hepatic iron overload. While the ill-posed field-to-source-inversion problem underlying QSM is conventionally assessed by the means of regularization techniques, we trained a fully convolutional deep neural network - DeepQSM - to directly invert the magnetic dipole kernel convolution. DeepQSM learned the physical forward problem using purely synthetic data and is capable of solving the ill-posed field-to-source inversion on in vivo MRI phase data. The magnetic susceptibility maps reconstructed by DeepQSM enable identification of deep brain substructures and provide information on their respective magnetic tissue properties. In summary, DeepQSM can invert the magnetic dipole kernel convolution and delivers robust solutions to this ill-posed problem.

The effects of variations in tissue microstructure from postmortem rat brain on the asymmetry of the water proton resonance

PURPOSE

This work was performed to investigate the effects of tissue microstructure from postmortem rat brain on the shape of the water proton spectrum.

METHODS

Perfusion-fixed, resected rat brains (N = 4) were imaged at 9.4T. 3D DTI and 3D echo-planar spectroscopic imaging (EPSI) data were acquired with 150 μm isotropic resolution. DTI data were acquired over 60 directions with b = 3000 s/mm² . Water spectra were produced from EPSI data acquired over 128 echoes, with 2.9 Hz spectral resolution. A voxel-wise metric reflecting spectral asymmetry about the peak of the resonance was computed and compared with orientation estimates from DTI data by fitting data with the susceptibility anisotropy model.

RESULTS

Asymmetric broadening of the water resonance was computed for mixed populations of grey and/or white matter as determined by thresholding the fractional anisotropy. Asymmetry was shown to be differentially affected by tract orientation relative to B₀ in high FA voxels, whereas low FA voxels exhibited little sensitivity. Anatomic structures in the hippocampus were also found to produce distinct changes in the water resonance.

CONCLUSION

Present results demonstrate that structural variations in tissue architecture cause characteristic, reproducible changes in the water resonance shape. This suggests that water spectra are sensitive to cytoarchitectural variations in brain tissue.

Signal compartments in ultra-high field multi-echo gradient echo MRI reflect underlying tissue microstructure in the brain

Gradient recalled echo magnetic resonance imaging (GRE-MRI) at ultra-high field holds great promise for new contrast mechanisms and delineation of putative tissue compartments that contribute to the multi-echo GRE-MRI signal may aid structural characterization. Several studies have adopted the three water-pool compartment model to study white matter brain regions, associating individual compartments with myelin, axonal and extracellular water. However, the number and identifiability of GRE-MRI signal compartments has not been fully explored. We undertook this task for human brain imaging data. Multiple echo time GRE-MRI data were acquired in five healthy participants, specific anatomical structures were segmented in each dataset (substantia nigra, caudate, insula, putamen, thalamus, fornix, internal capsule, corpus callosum and cerebrospinal fluid), and the signal fitted with models comprising one to six signal compartments using a complex-valued plane wave formulation. Information criteria and cluster analysis methods were used to ascertain the number of distinct compartments within the signal from each structure and to determine their respective frequency shifts. We identified five principal signal compartments with different relative contributions to each structure's signal. Voxel-based maps of the volume fraction of each of these compartments were generated and demonstrated spatial correlation with brain anatomy.