Artificial neural networks for magnetic resonance elastography stiffness estimation in inhomogeneous materials

Evaluation of MR Elastography as a Noninvasive Diagnostic Test for Spontaneous Intracranial Hypotension

BACKGROUND AND PURPOSE

Spontaneous intracranial hypotension is a condition resulting from a leak of CSF from the spinal canal arising independent of a medical procedure. Spontaneous intracranial hypotension can present with normal brain MR imaging findings and nonspecific symptoms, leading to the underdiagnosis in some patients and unnecessary invasive myelography in others who are found not to have the condition. Given the likelihood that spontaneous intracranial hypotension alters intracranial biomechanics, the goal of this study was to evaluate MR elastography as a potential noninvasive test to diagnose the condition.

MATERIALS AND METHODS

We performed MR elastography in 15 patients with confirmed spontaneous intracranial hypotension from September 2022 to April 2023. Age, sex, symptom duration, and brain MR imaging Bern score were collected. MR elastography data were used to compute stiffness and damping ratio maps, and voxelwise modeling was performed to detect clusters of significant differences in mechanical properties between patients with spontaneous intracranial hypotension and healthy control participants. To evaluate diagnostic accuracy, we summarized each examination by 2 spatial pattern scores (one each for stiffness and damping ratio) and evaluated group-wise discrimination by receiver operating characteristic curve analysis.

RESULTS

Patients with spontaneous intracranial hypotension exhibited significant differences in both stiffness and damping ratio (false discovery rate-corrected, Q < 0.05). Pattern analysis discriminated patients with spontaneous intracranial hypotension from healthy controls with an area under the curve of 0.97 overall, and the area under the curve was 0.97 in those without MR imaging findings of spontaneous intracranial hypotension.

CONCLUSIONS

Results from this pilot study demonstrate MR elastography as a potential imaging biomarker and a noninvasive method for diagnosing spontaneous intracranial hypotension, including patients with normal brain MR imaging findings.

A Comparative Study of Three Systems for Liver Magnetic Resonance Elastography

BACKGROUND

Different MR elastography (MRE) systems may produce different stiffness measurements, making direct comparison difficult in multi-center investigations.

PURPOSE

To assess the repeatability and reproducibility of liver stiffness measured by three typical MRE systems.

STUDY TYPE

Prospective.

POPULATION/PHANTOMS

Thirty volunteers without liver disease history (20 males, aged 21-28)/5 gel phantoms.

FIELD STRENGTH/SEQUENCE

3.0 T United Imaging Healthcare (UIH), 1.5 T Siemens Healthcare, 3.0 T General Electric Healthcare (GE)/Echo planar imaging-based MRE sequence.

ASSESSMENT

Wave images of volunteers and phantoms were acquired by three MRE systems. Tissue stiffness was evaluated by two observers, while phantom stiffness was assessed automatically by code. The reproducibility across three MRE systems was quantified based on the mean stiffness of each volunteer and phantom.

STATISTICAL TESTS

Intraclass correlation coefficients (ICC), coefficients of variation (CV), and Bland-Altman analyses were used to assess the interobserver reproducibility, the interscan repeatability, and the intersystem reproducibility. Paired t-tests were performed to assess the interobserver and interscan variation. Friedman tests with Dunn's multiple comparison correction were performed to assess the intersystem variation. P values less than 0.05 indicated significant difference.

RESULTS

The reproducibility of stiffness measured by the two observers demonstrated consistency with ICC > 0.92, CV < 4.32%, Mean bias < 2.23%, and P > 0.06. The repeatability of measurements obtained using the electromagnetic system for the liver revealed ICC > 0.96, CV < 3.86%, Mean bias < 0.19%, P > 0.90. When considering the range of reproducibility across the three systems for liver evaluations, results ranged with ICCs from 0.70 to 0.87, CVs from 6.46% to 10.99%, and Mean biases between 1.89% and 6.30%. Phantom studies showed similar results. The values of measured stiffness differed across all three systems significantly.

DATA CONCLUSION

Liver stiffness values measured from different MRE systems can be different, but the measurements across the three MRE systems produced consistent results with excellent reproducibility.

EVIDENCE LEVEL

1 TECHNICAL EFFICACY: Stage 2.

Mechanics of Small-Scale Spherical Inclusions Using Nonlocal Poroelasticity Integrated with Light Gradient Boosting Machine

Detecting inclusions in materials at small scales is of high importance to ensure the quality, structural integrity and performance efficiency of microelectromechanical machines and products. Ultrasound waves are commonly used as a non-destructive method to find inclusions or structural flaws in a material. Mathematical continuum models can be used to enable ultrasound techniques to provide quantitative information about the change in the mechanical properties due to the presence of inclusions. In this paper, a nonlocal size-dependent poroelasticity model integrated with machine learning is developed for the description of the mechanical behaviour of spherical inclusions under uniform radial compression. The scale effects on fluid pressure and radial displacement are captured using Eringen's theory of nonlocality. The conservation of mass law is utilised for both the solid matrix and fluid content of the poroelastic material to derive the storage equation. The governing differential equations are derived by decoupling the equilibrium equation and effective stress-strain relations in the spherical coordinate system. An accurate numerical solution is obtained using the Galerkin discretisation technique and a precise integration method. A Dormand-Prince solution is also developed for comparison purposes. A light gradient boosting machine learning model in conjunction with the nonlocal model is used to extract the pattern of changes in the mechanical response of the poroelastic inclusion. The optimised hyperparameters are calculated by a grid search cross validation. The modelling estimation power is enhanced by considering nonlocal effects and applying machine learning processes, facilitating the detection of ultrasmall inclusions within a poroelastic medium at micro/nanoscales.

Clinical application of magnetic resonance elastography in pediatric neurological disorders

Magnetic resonance elastography is a relatively new, rapidly evolving quantitative magnetic resonance imaging technique which can be used for mapping the viscoelastic mechanical properties of soft tissues. MR elastography measurements are akin to manual palpation but with the advantages of both being quantitative and being useful for regions which are not available for palpation, such as the human brain. MR elastography is noninvasive, well tolerated, and complements standard radiological and histopathological studies by providing in vivo measurements that reflect tissue microstructural integrity. While brain MR elastography studies in adults are becoming frequent, published studies on the utility of MR elastography in children are sparse. In this review, we have summarized the major scientific principles and recent clinical applications of brain MR elastography in diagnostic neuroscience and discuss avenues for impact in assessing the pediatric brain.

MR Elastography With Optimization-Based Phase Unwrapping and Traveling Wave Expansion-Based Neural Network (TWENN)

Magnetic Resonance Elastography (MRE) can characterize biomechanical properties of soft tissue for disease diagnosis and treatment planning. However, complicated wavefields acquired from MRE coupled with noise pose challenges for accurate displacement extraction and modulus estimation. Using optimization-based displacement extraction and Traveling Wave Expansion-based Neural Network (TWENN) modulus estimation, we propose a new pipeline for processing MRE images. An objective function with Dual Data Consistency (Dual-DC) has been used to ensure accurate phase unwrapping and displacement extraction. For the estimation of complex wavenumbers, a complex-valued neural network with displacement covariance as an input has been developed. A model of traveling wave expansion is used to generate training datasets for the network with varying levels of noise. The complex shear modulus map is obtained through fusion of multifrequency and multidirectional data. Validation using brain and liver simulation images demonstrates the practical value of the proposed pipeline, which can estimate the biomechanical properties with minimal root-mean-square errors when compared to state-of-the-art methods. Applications of the proposed method for processing MRE images of phantom, brain, and liver reveal clear anatomical features, robustness to noise, and good generalizability of the pipeline.

Reconstructing the Spatial Distribution of the Relative Shear Modulus in Quasi-static Ultrasound Elastography: Plane Stress Analysis

Quasi-static ultrasound elastography (QSUE) is an imaging technique that mainly provides axial strain maps of tissues when the latter are subjected to compression. In this article, a method for reconstructing the relative shear modulus distribution within a linear elastic and isotropic medium, in QSUE, is introduced. More specifically, the plane stress inverse problem is considered. The proposed method is based on the variational formulation of the equilibrium equations and on the choice of adapted discretization spaces, and only requires displacement fields in the analyzed media to be determined. Results from plane stress and 3-D numerical simulations, as well as from phantom experiments, showed that the method is able to reconstruct the different regions within a medium, with shear modulus contrasts that unambiguously reveal whether inclusions are stiffer or softer than the surrounding material. More specifically, for the plane stress simulations, inclusion-to-background modulus ratios were found to be very accurately estimated, with an error lower than 3%. For the 3-D simulations, for which the plane stress conditions are no longer satisfied, these ratios were, as expected, less accurate, with an error that remained lower than 10% for two of the three cases analyzed but was around 34% for the last case. Concerning the phantom experiments, a comparison with a shear wave elastography technique from a clinical ultrasound scanner was also made. Overall, the inclusion-to-background shear modulus ratios obtained with our approach were found to be closer to those given by the phantom manufacturer than the ratios provided by the clinical system.

Reverberant magnetic resonance elastographic imaging using a single mechanical driver

Reverberant elastography provides fast and robust estimates of shear modulus; however, its reliance on multiple mechanical drivers hampers clinical utility. In this work, we hypothesize that for constrained organs such as the brain, reverberant elastography can produce accurate magnetic resonance elastograms with a single mechanical driver. To corroborate this hypothesis, we performed studies on healthy volunteers (n= 3); and a constrained calibrated brain phantom containing spherical inclusions with diameters ranging from 4-18 mm. In both studies (i.e. phantom and clinical), imaging was performed at frequencies of 50 and 70 Hz. We used the accuracy and contrast-to-noise ratio performance metrics to evaluate reverberant elastograms relative to those computed using the established subzone inversion method. Errors incurred in reverberant elastograms varied from 1.3% to 16.6% when imaging at 50 Hz and 3.1% and 16.8% when imaging at 70 Hz. In contrast, errors incurred in subzone elastograms ranged from 1.9% to 13% at 50 Hz and 3.6% to 14.9% at 70 Hz. The contrast-to-noise ratio of reverberant elastograms ranged from 63.1 to 73 dB compared to 65 to 66.2 dB for subzone elastograms. The average global brain shear modulus estimated from reverberant and subzone elastograms was 2.36 ± 0.07 kPa and 2.38 ± 0.11 kPa, respectively, when imaging at 50 Hz and 2.70 ± 0.20 kPa and 2.89 ± 0.60 kPa respectively, when imaging at 70 Hz. The results of this investigation demonstrate that reverberant elastography can produce accurate, high-quality elastograms of the brain with a single mechanical driver.

OSCILLATE: A low-rank approach for accelerated magnetic resonance elastography

PURPOSE

MR elastography (MRE) is a technique to characterize brain mechanical properties in vivo. Due to the need to capture tissue deformation in multiple directions over time, MRE is an inherently long acquisition, which limits achievable resolution and use in challenging populations. The purpose of this work is to develop a method for accelerating MRE acquisition by using low-rank image reconstruction to exploit inherent spatiotemporal correlations in MRE data.

THEORY AND METHODS

The proposed MRE sampling and reconstruction method, OSCILLATE (Observing Spatiotemporal Correlations for Imaging with Low-rank Leveraged Acceleration in Turbo Elastography), involves alternating which k-space points are sampled between each repetition by a reduction factor, ROSC. Using a predetermined temporal basis from a low-resolution navigator in a joint low-rank image reconstruction, all images can be accurately reconstructed from a reduced amount of k-space data.

RESULTS

Decomposition of MRE displacement data demonstrated that, on average, 96.1% of all energy from an MRE dataset is captured at rank L = 12 (reduced from a full rank of 24). Retrospectively undersampling data with ROSC  = 2 and reconstructing at low-rank (L = 12) yields highly accurate stiffness maps with voxel-wise error of 5.8% ± 0.7%. Prospectively undersampled data at ROSC  = 2 were successfully reconstructed without loss of material property map fidelity, with average global stiffness error of 1.0% ± 0.7% compared to fully sampled data.

CONCLUSIONS

OSCILLATE produces whole-brain MRE data at 2 mm isotropic resolution in 1 min 48 s.

Numerical simulation of wave propagation through interfaces using the extended finite element method for magnetic resonance elastography

Magnetic resonance elastography (MRE) is an elasticity imaging technique for quantitatively assessing the stiffness of human tissues. In MRE, finite element method (FEM) is widely used for modeling wave propagation and stiffness reconstruction. However, in front of inclusions with complex interfaces, FEM can become burdensome in terms of the model partition and computationally expensive. In this work, we implement a formulation of FEM, known as the eXtended finite element method (XFEM), which is a method used for modeling discontinuity like crack and heterogeneity. Using a level-set method, it makes the interface independent of the mesh, thus relieving the meshing efforts. We investigate this method in two studies: wave propagation across an oblique linear interface and stiffness reconstruction of a random-shape inclusion. In the first study, numerical results by XFEM and FEM models revealing the wave conversion rules at linear interface are presented and successfully compared to the theoretical predictions. The second study, investigated in a pseudo-practical application, demonstrates further the applicability of XFEM in MRE and the convenience, accuracy, and speed of XFEM with respect to FEM. XFEM can be regarded as a promising alternative to FEM for inclusion modeling in MRE.

Impact of material homogeneity assumption on cortical stiffness estimates by MR elastography

PURPOSE

Inversion algorithms used to convert acquired MR elastography wave data into material property estimates often assume that the underlying materials are locally homogeneous. Here we evaluate the impact of that assumption on stiffness estimates in gray-matter regions of interest in brain MR elastography.

METHODS

We describe an updated neural network inversion framework using finite-difference model-derived data to train convolutional neural network inversion algorithms. Neural network inversions trained on homogeneous simulations (homogeneous learned inversions [HLIs]) or inhomogeneous simulations (inhomogeneous learned inversions [ILIs]) are generated with a variety of kernel sizes. These inversions are evaluated in a brain MR elastography simulation experiment and in vivo in a test-retest repeatability experiment including 10 healthy volunteers.

RESULTS

In simulation and in vivo, HLI and ILI with small kernels produce similar results. As kernel size increases, the assumption of homogeneity has a larger effect, and HLI and ILI stiffness estimates show larger differences. At each inversion's optimal kernel size in simulation (7 × 7 × 7 for HLI, 11 × 11 × 11 for ILI), ILI is more sensitive to true changes in stiffness in gray-matter regions of interest in simulation. In vivo, there is no difference in the region-level repeatability of stiffness estimates between the inversions, although ILI appears to better maintain the stiffness map structure as kernel size increases, while decreasing the spatial variance in stiffness estimates.

CONCLUSIONS

This study suggests that inhomogeneous inversions provide small but significant benefits even when large stiffness gradients are absent.

In vivo stiffness of multiple sclerosis lesions is similar to that of normal-appearing white matter

In 1868, French neurologist Jean-Martin Charcot coined the term multiple sclerosis (MS) after his observation that numerous white matter (WM) glial scars felt like sclerotic tissue. Nowadays, magnetic resonance elastography (MRE) can generate images with contrast of stiffness (CS) in soft in vivo tissues and may therefore be sensitive to MS lesions, provided that sclerosis is indeed a mechanical signature of this disease. We analyzed CS in a total of 147 lesions in patients with relapsing-remitting MS, compared with control regions in contralateral brain regions, and phantom data as well as performed numerical simulations to determine the delineation limits of multifrequency MRE (20 - 40 Hz) in MS. MRE analysis of simulated waves revealed a delineation limit of approximately 10% CS for detecting 9-mm lesions (mean size in our patient population). Due to inversion bias, this limit is reached when true CS is -11% for soft and 35% for stiff lesions. In vivo MRE identified 35 stiffer lesions and 17 softer lesions compared with surrounding WM (mean stiffness: 934±82 Pa). However, a similar pattern was found in the contralateral brain, suggesting that the range of stiffness changes in WM lesions due to MS is within the normal range of WM variability and normal heterogeneity-related CS. Consequently, Charcot's original intuition that MS is a focal sclerotic disease can neither be dismissed nor confirmed by in vivo MRE. However, the observation that MS lesions do not markedly differ in stiffness from surrounding brain tissue suggests that marked tissue sclerosis is not a mechanical signature of MS. STATEMENT OF SIGNIFICANCE: Multiple sclerosis (MS) was named by J.M. Charcot after the sclerotic changes in brain tissue he found in post-mortem autopsies. Since then, nothing has been revealed about the actual stiffening of MS lesions in vivo. Studying the viscoelastic properties of plaques in their natural environment is a major challenge that can only be overcome by MR elastography (MRE). Therefore, we used multifrequency MRE to answer the question whether MS lesions in patients with a relapsing-remitting disease course are mechanically different than surrounding tissue. Our findings suggest that the range of stiffness changes in white matter lesions due to MS is within the normal range of white matter variability and in vivo tissue sclerosis might not be a mechanical signature of MS.

MR elastography inversion by compressive recovery

Direct inversion (DI) derives tissue shear modulus by inverting the Helmholtz equation. However, conventional DI is sensitive to data quality due to the ill-posed nature of Helmholtz inversion and thus providing reliable stiffness estimation can be challenging. This becomes more problematic in the case of estimating shear stiffness of the lung in which the low tissue density and short T2* result in considerably low signal-to-noise ratio during lung MRE. In the present study, we propose to perform MRE inversion by compressive recovery (MICRo). Such a technique aims to improve the numerical stability and the robustness to data noise of Helmholtz inversion by using prior knowledge on data noise and transform sparsity of the stiffness map. The developed inversion strategy was first validated in simulated phantoms with known stiffness. Next, MICRo was compared to the standard clinical multi-modal DI (MMDI) method forin vivoliver MRE in healthy subjects and patients with different stages of liver fibrosis. After establishing the accuracy of MICRo, we demonstrated the robustness of the proposed technique against data noise in lung MRE with healthy subjects. In simulated phantoms with single-directional or multi-directional waves, MICRo outperformed DI with Romano filter or Savitsky and Golay filter, especially when the stiffness and/or noise level was high. In hepatic MRE application, agreement was observed between MICRo and MMDI. Measuringin vivolung stiffness, MICRo demonstrated its advantages over filtered DI by yielding stable stiffness estimation at both residual volume and total lung capacity. These preliminary results demonstrate the potential value of the proposed technique and also warrant further investigation in a larger clinical population.

Harnessing brain waves: a review of brain magnetic resonance elastography for clinicians and scientists entering the field

Brain magnetic resonance elastography (MRE) is an imaging technique capable of accurately and non-invasively measuring the mechanical properties of the living human brain. Recent studies have shown that MRE has potential to provide clinically useful information in patients with intracranial tumors, demyelinating disease, neurodegenerative disease, elevated intracranial pressure, and altered functional states. The objectives of this review are: (1) to give a general overview of the types of measurements that have been obtained with brain MRE in patient populations, (2) to survey the tools currently being used to make these measurements possible, and (3) to highlight brain MRE-based quantitative biomarkers that have the highest potential of being adopted into clinical use within the next 5 to 10 years. The specifics of MRE methodology strategies are described, from wave generation to material parameter estimations. The potential clinical role of MRE for characterizing and planning surgical resection of intracranial tumors and assessing diffuse changes in brain stiffness resulting from diffuse neurological diseases and altered intracranial pressure are described. In addition, the emerging technique of functional MRE, the role of artificial intelligence in MRE, and promising applications of MRE in general neuroscience research are presented.

MR elastography: Principles, guidelines, and terminology

Magnetic resonance elastography (MRE) is a phase contrast-based MRI technique that can measure displacement due to propagating mechanical waves, from which material properties such as shear modulus can be calculated. Magnetic resonance elastography can be thought of as quantitative, noninvasive palpation. It is increasing in clinical importance, has become widespread in the diagnosis and staging of liver fibrosis, and additional clinical applications are being explored. However, publications have reported MRE results using many different parameters, acquisition techniques, processing methods, and varied nomenclature. The diversity of terminology can lead to confusion (particularly among clinicians) about the meaning of and interpretation of MRE results. This paper was written by the MRE Guidelines Committee, a group formalized at the first meeting of the ISMRM MRE Study Group, to clarify and move toward standardization of MRE nomenclature. The purpose of this paper is to (1) explain MRE terminology and concepts to those not familiar with them, (2) define "good practices" for practitioners of MRE, and (3) identify opportunities to standardize terminology, to avoid confusion.