LARO: Learned acquisition and reconstruction optimization to accelerate quantitative susceptibility mapping.
Ontology highlight
ABSTRACT:
SUBMITTER: Zhang J
PROVIDER: S-EPMC10021353 | biostudies-literature | 2023 Mar
REPOSITORIES: biostudies-literature
Publications
LARO: Learned acquisition and reconstruction optimization to accelerate quantitative susceptibility mapping.
Zhang Jinwei J Spincemaille Pascal P Zhang Hang H Nguyen Thanh D TD Li Chao C Li Jiahao J Kovanlikaya Ilhami I Sabuncu Mert R MR Wang Yi Y
NeuroImage 20230117
Quantitative susceptibility mapping (QSM) involves acquisition and reconstruction of a series of images at multi-echo time points to estimate tissue field, which prolongs scan time and requires specific reconstruction technique. In this paper, we present our new framework, called Learned Acquisition and Reconstruction Optimization (LARO), which aims to accelerate the multi-echo gradient echo (mGRE) pulse sequence for QSM. Our approach involves optimizing a Cartesian multi-echo k-space sampling p ...[more]
Similar Datasets
Project description:Quantitative Susceptibility Mapping (QSM) estimates tissue magnetic susceptibility distributions from Magnetic Resonance (MR) phase measurements by solving an ill-posed dipole inversion problem. Conventional single orientation QSM methods usually employ regularization strategies to stabilize such inversion, but may suffer from streaking artifacts or over-smoothing. Multiple orientation QSM such as calculation of susceptibility through multiple orientation sampling (COSMOS) can give well-conditioned inversion and an artifact free solution but has expensive acquisition costs. On the other hand, Convolutional Neural Networks (CNN) show great potential for medical image reconstruction, albeit often with limited interpretability. Here, we present a Learned Proximal Convolutional Neural Network (LP-CNN) for solving the ill-posed QSM dipole inversion problem in an iterative proximal gradient descent fashion. This approach combines the strengths of data-driven restoration priors and the clear interpretability of iterative solvers that can take into account the physical model of dipole convolution. During training, our LP-CNN learns an implicit regularizer via its proximal, enabling the decoupling between the forward operator and the data-driven parameters in the reconstruction algorithm. More importantly, this framework is believed to be the first deep learning QSM approach that can naturally handle an arbitrary number of phase input measurements without the need for any ad-hoc rotation or re-training. We demonstrate that the LP-CNN provides state-of-the-art reconstruction results compared to both traditional and deep learning methods while allowing for more flexibility in the reconstruction process.
| S-EPMC7647124 | biostudies-literature
Project description:PurposeQuantitative Susceptibility Mapping (QSM) is usually performed by minimizing a functional with data fidelity and regularization terms. A weighting parameter controls the balance between these terms. There is a need for techniques to find the proper balance that avoids artifact propagation and loss of details. Finding the point of maximum curvature in the L-curve is a popular choice, although it is slow, often unreliable when using variational penalties, and has a tendency to yield overregularized results.MethodsWe propose 2 alternative approaches to control the balance between the data fidelity and regularization terms: 1) searching for an inflection point in the log-log domain of the L-curve, and 2) comparing frequency components of QSM reconstructions. We compare these methods against the conventional L-curve and U-curve approaches.ResultsOur methods achieve predicted parameters that are better correlated with RMS error, high-frequency error norm, and structural similarity metric-based parameter optimizations than those obtained with traditional methods. The inflection point yields less overregularization and lower errors than traditional alternatives. The frequency analysis yields more visually appealing results, although with larger RMS error.ConclusionOur methods provide a robust parameter optimization framework for variational penalties in QSM reconstruction. The L-curve-based zero-curvature search produced almost optimal results for typical QSM acquisition settings. The frequency analysis method may use a 1.5 to 2.0 correction factor to apply it as a stand-alone method for a wider range of signal-to-noise-ratio settings. This approach may also benefit from fast search algorithms such as the binary search to speed up the process.
| S-EPMC7722179 | biostudies-literature
Project description:Quantitative susceptibility mapping (QSM) estimates the underlying tissue magnetic susceptibility from MRI gradient-echo phase signal and typically requires several processing steps. These steps involve phase unwrapping, brain volume extraction, background phase removal and solving an ill-posed inverse problem relating the tissue phase to the underlying susceptibility distribution. The resulting susceptibility map is known to suffer from inaccuracy near the edges of the brain tissues, in part due to imperfect brain extraction, edge erosion of the brain tissue and the lack of phase measurement outside the brain. This inaccuracy has thus hindered the application of QSM for measuring susceptibility of tissues near the brain edges, e.g., quantifying cortical layers and generating superficial venography. To address these challenges, we propose a learning-based QSM reconstruction method that directly estimates the magnetic susceptibility from total phase images without the need for brain extraction and background phase removal, referred to as autoQSM. The neural network has a modified U-net structure and is trained using QSM maps computed by a two-step QSM method. 209 healthy subjects with ages ranging from 11 to 82 years were employed for patch-wise network training. The network was validated on data dissimilar to the training data, e.g., in vivo mouse brain data and brains with lesions, which suggests that the network generalized and learned the underlying mathematical relationship between magnetic field perturbation and magnetic susceptibility. Quantitative and qualitative comparisons were performed between autoQSM and other two-step QSM methods. AutoQSM was able to recover magnetic susceptibility of anatomical structures near the edges of the brain including the veins covering the cortical surface, spinal cord and nerve tracts near the mouse brain boundaries. The advantages of high-quality maps, no need for brain volume extraction, and high reconstruction speed demonstrate autoQSM's potential for future applications.
| S-EPMC6819263 | biostudies-literature
Project description:PurposeTo enable free-breathing and high isotropic resolution liver quantitative susceptibility mapping (QSM) using 3D multi-echo UTE cones acquisition and respiratory motion-resolved image reconstruction.MethodsUsing 3D multi-echo UTE cones MRI, a respiratory motion was estimated from the k-space center of the imaging data. After sorting the k-space data with estimated motion, respiratory motion state-resolved reconstruction was performed for multi-echo data followed by nonlinear least-squares fitting for proton density fat fraction (PDFF), R2*$$ {\mathrm{R}}_2^{\ast } $$ , and fat-corrected B0 field maps. PDFF and B0 field maps were subsequently used for QSM reconstruction. The proposed method was compared with motion-averaged (gridding) reconstruction and conventional 3D multi-echo Cartesian MRI in moving gadolinium phantom and in vivo studies. Region of interest (ROI)-based linear regression analysis was performed on these methods to investigate correlations between gadolinium concentration and QSM in the phantom study and between R2*$$ {\mathrm{R}}_2^{\ast } $$ and QSM in in vivo study.ResultsCones with motion-resolved reconstruction showed sharper image quality compared to motion-averaged reconstruction with a substantial reduction of motion artifacts in both moving phantom and in vivo studies. For ROI-based linear regression analysis of the phantom study, susceptibility values from cones with motion-resolved reconstruction ( QSMppm$$ {\mathrm{QSM}}_{\mathrm{ppm}} $$ = 0.31 ×gadoliniummM+$$ \times {\mathrm{gadolinium}}_{\mathrm{mM}}+ $$ 0.05, R2$$ {R}^2 $$ = 0.999) and Cartesian without motion ( QSMppm$$ {\mathrm{QSM}}_{\mathrm{ppm}} $$ = 0.32 ×gadoliniummM+$$ \times {\mathrm{gadolinium}}_{\mathrm{mM}}+ $$ 0.04, R2$$ {R}^2 $$ = 1.000) showed linear relationships with gadolinium concentrations and showed good agreement with each other. For in vivo, motion-resolved reconstruction showed higher goodness of fit ( QSMppm$$ {\mathrm{QSM}}_{\mathrm{ppm}} $$ = 0.00261 ×R2s-1*-$$ \times {\mathrm{R}}_{2_{{\mathrm{s}}^{-1}}}^{\ast }- $$ 0.524, R2$$ {R}^2 $$ = 0.977) compared to motion-averaged reconstruction ( QSMppm$$ {\mathrm{QSM}}_{\mathrm{ppm}} $$ = 0.0021 ×R2s-1*-$$ \times {\mathrm{R}}_{2_{{\mathrm{s}}^{-1}}}^{\ast }- $$ 0.572, R2$$ {R}^2 $$ = 0.723) in ROI-based linear regression analysis between R2*$$ {\mathrm{R}}_2^{\ast } $$ and QSM.ConclusionFeasibility of free-breathing liver QSM was demonstrated with motion-resolved 3D multi-echo UTE cones MRI, achieving high isotropic resolution currently unachievable in conventional Cartesian MRI.
| S-EPMC10529485 | biostudies-literature
Project description:The reconstruction of MR quantitative susceptibility mapping (QSM) from local phase measurements is an ill posed inverse problem and different regularization strategies incorporating a priori information extracted from magnitude and phase images have been proposed. However, the anatomy observed in magnitude and phase images does not always coincide spatially with that in susceptibility maps, which could give erroneous estimation in the reconstructed susceptibility map. In this paper, we develop a structural feature based collaborative reconstruction (SFCR) method for QSM including both magnitude and susceptibility based information. The SFCR algorithm is composed of two consecutive steps corresponding to complementary reconstruction models, each with a structural feature based l 1 norm constraint and a voxel fidelity based l 2 norm constraint, which allows both the structure edges and tiny features to be recovered, whereas the noise and artifacts could be reduced. In the M-step, the initial susceptibility map is reconstructed by employing a k -space based compressed sensing model incorporating magnitude prior. In the S-step, the susceptibility map is fitted in spatial domain using weighted constraints derived from the initial susceptibility map from the M-step. Simulations and in vivo human experiments at 7T MRI show that the SFCR method provides high quality susceptibility maps with improved RMSE and MSSIM. Finally, the susceptibility values of deep gray matter are analyzed in multiple head positions, with the supine position most approximate to the gold standard COSMOS result.
| S-EPMC5495149 | biostudies-literature
Project description:Quantitative Susceptibility Mapping (QSM) is an established Magnetic Resonance Imaging (MRI) technique with high potential in brain iron studies associated to several neurodegenerative diseases. Unlike other MRI techniques, QSM relies on phase images to estimate tissue's relative susceptibility, therefore requiring a reliable phase data. Phase images from a multi-channel acquisition should be reconstructed in a proper way. On this work it was compared the performance of combination of phase matching algorithms (MCPC3D-S and VRC) and phase combination methods based on a complex weighted sum of phases, considering the magnitude at different powers (k = 0 to 4) as the weighting factor. These reconstruction methods were applied in two datasets: a simulated brain dataset for a 4-coil array and data of 22 postmortem subjects acquired at a 7T scanner using a 32 channels coil. For the simulated dataset, differences between the ground truth and the Root Mean Squared Error (RMSE) were evaluated. For both simulated and postmortem data, the mean (MS) and standard deviation (SD) of susceptibility values of five deep gray matter regions were calculated. For the postmortem subjects, MS and SD were statistically compared across all subjects. A qualitative analysis indicated no differences between methods, except for the Adaptive approach on postmortem data, which showed intense artifacts. In the 20% noise level case, the simulated data showed increased noise in central regions. Quantitative analysis showed that both MS and SD were not statistically different when comparing k = 1 and k = 2 on postmortem brain images, however visual inspection showed some boundaries artifacts on k = 2. Furthermore, the RMSE decreased (on regions near the coils) and increased (on central regions and on overall QSM) with increasing k. In conclusion, for reconstruction of phase images from multiple coils with no reference available, alternative methods are needed. In this study it was found that overall, the phase combination with k = 1 is preferred over other powers of k.
| S-EPMC10062192 | biostudies-literature
Project description:To apply quantitative susceptibility mapping (QSM) in the basal ganglia of patients with multiple sclerosis (MS) and relate the findings to R2* mapping with regard to the sensitivity for clinical and morphologic measures of disease severity.The local ethics committee approved this study, and all subjects gave written informed consent. Sixty-eight patients (26 with clinically isolated syndrome, 42 with relapsing-remitting MS) and 23 control subjects underwent 3-T magnetic resonance (MR) imaging. Susceptibility and R2* maps were reconstructed from the same three-dimensional multiecho spoiled gradient-echo sequence. Mean susceptibilities and R2* rates were measured in the basal ganglia and were compared between different phenotypes of the disease (clinically isolated syndrome, MS) and the control subjects by using analysis of variance, and regressing analysis was used to identify independent predictors.Compared with control subjects, patients with MS and clinically isolated syndrome had increased (more paramagnetic) magnetic susceptibilities in the basal ganglia. R2* mapping proved less sensitive than QSM regarding group differences. The strongest predictor of magnetic susceptibility was age. Susceptibilities were higher with increasing neurologic deficits (r = 0.34, P < .01) and lower with normalized volumes of gray matter (r = -0.35, P < .005) and the cortex (r = -0.35, P < .005).QSM provides superior sensitivity over R2* mapping in the detection of MS-related tissue changes in the basal ganglia. With QSM but not with R2* mapping, changes were already observed in patients with clinically isolated syndrome, which suggests that QSM can serve as a sensitive measure at the earliest stage of the disease.
| S-EPMC3632806 | biostudies-other
Project description:BackgroundAccurate measurement of the liver iron concentration (LIC) is needed to guide iron-chelating therapy for patients with transfusional iron overload. In this work, we investigate the feasibility of automated quantitative susceptibility mapping (QSM) to measure the LIC.PurposeTo develop a rapid, robust, and automated liver QSM for clinical practice.Study typeProspective.Population13 healthy subjects and 22 patients.Field strength/sequences1.5 T and 3 T/3D multiecho gradient-recalled echo (GRE) sequence.AssessmentData were acquired using a 3D GRE sequence with an out-of-phase echo spacing with respect to each other. All odd echoes that were in-phase (IP) were used to initialize the fat-water separation and field estimation (T2 *-IDEAL) before performing QSM. Liver QSM was generated through an automated pipeline without manual intervention. This IP echo-based initialization method was compared with an existing graph cuts initialization method (simultaneous phase unwrapping and removal of chemical shift, SPURS) in healthy subjects (n = 5). Reproducibility was assessed over four scanners at two field strengths from two manufacturers using healthy subjects (n = 8). Clinical feasibility was evaluated in patients (n = 22).Statistical testsIP and SPURS initialization methods in both healthy subjects and patients were compared using paired t-test and linear regression analysis to assess processing time and region of interest (ROI) measurements. Reproducibility of QSM, R2 *, and proton density fat fraction (PDFF) among the four different scanners was assessed using linear regression, Bland-Altman analysis, and the intraclass correlation coefficient (ICC).ResultsLiver QSM using the IP method was found to be ~5.5 times faster than SPURS (P < 0.05) in initializing T2 *-IDEAL with similar outputs. Liver QSM using the IP method were reproducibly generated in all four scanners (average coefficient of determination 0.95, average slope 0.90, average bias 0.002 ppm, 95% limits of agreement between -0.06 to 0.07 ppm, ICC 0.97).Data conclusionUse of IP echo-based initialization enables robust water/fat separation and field estimation for automated, rapid, and reproducible liver QSM for clinical applications.Level of evidence1 Technical Efficacy: Stage 2 J. Magn. Reson. Imaging 2019;50:725-732.
| S-EPMC6929208 | biostudies-literature
Project description:Various regularization methods have been proposed for single-orientation quantitative susceptibility mapping (QSM), which is an ill-posed magnetic field to susceptibility source inverse problem. Noise amplification, a major issue in inverse problems, manifests as streaking artifacts and quantification errors in QSM and has not been comparatively evaluated in these algorithms. In this paper, various QSM methods were systematically categorized for noise analysis. Six representative QSM methods were selected from four categories: two non-Bayesian methods with alteration or approximation of the dipole kernel to overcome the ill conditioning; four Bayesian methods using a general mathematical prior or a specific physical structure prior to select a unique solution, and using a data fidelity term with or without noise weighting. The effects of noise in these QSM methods were evaluated by reconstruction errors in simulation and image quality in 50 consecutive human subjects. Bayesian QSM methods with noise weighting consistently reduced root mean squared errors in numerical simulations and increased image quality scores in the human brain images, when compared to non-Bayesian methods and to corresponding Bayesian methods without noise weighting (p ≤ 0.001). In summary, noise effects in QSM can be reduced using Bayesian methods with proper noise weighting.
| S-EPMC5553691 | biostudies-other
Project description:PurposeTo develop an automated adaptive preconditioner for QSM reconstruction with improved susceptibility quantification accuracy and increased image quality.Theory and methodsThe total field was used to rapidly produce an approximate susceptibility map, which was then averaged and trended over R2∗ binning to generate a spatially varying distribution of preconditioning values. This automated adaptive preconditioner was used to reconstruct QSM via total field inversion and was compared with its empirical counterparts in a numerical simulation, a brain experiment with 5 healthy subjects and 5 patients with intracerebral hemorrhage, and a cardiac experiment with 3 healthy subjects.ResultsAmong evaluated preconditioners, the automated adaptive preconditioner achieved the fastest convergence in reducing the RMSE of the QSM in the simulation, suppressed hemorrhage-associated artifacts while preserving surrounding brain tissue contrasts, and provided cardiac chamber oxygenation values consistent with those reported in the literature.ConclusionAn automated adaptive preconditioner allows high-quality QSM from the total field in imaging various anatomies with dynamic susceptibility ranges.
| S-EPMC6778703 | biostudies-literature