Bernhard Strasser
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Bernhard Strasser has a total of 283 co-authors. Here are the top co-authors by number of shared publications:
- Wolfgang Bogner
High Field MR Center, Department of Biomedical Imaging and Image‐guided Therapy Medical University of Vienna Vienna Austria
41 shared publications - Siegfried Trattnig
High Field MR Center, Department of Biomedical Imaging and Image‐guided Therapy Medical University of Vienna Vienna Austria
23 shared publications - Gilbert Hangel
Department of Neurosurgery, Medical University of Vienna, Vienna, Austria
22 shared publications - Lukas Hingerl
Department of Biomedical Imaging and Image-guided Therapy, Radiology and Nuclear Medicine, Medical University of Vienna, Vienna, Austria
22 shared publications - Stanislav Motyka
High Field MR Center, Department of Biomedical Imaging and Image‐guided Therapy Medical University of Vienna Vienna Austria
12 shared publications - Fabian Niess
High-field MR Center, Department of Biomedical Imaging and Image-guided Therapy, Medical University of Vienna, Vienna, Austria
10 shared publications
Latest Publications
- Open Access
- 17 November 2026
- Published by Wiley in NMR in Biomedicine
- Vol. 39 (12)
Synthetic Data in MR Spectroscopy: Current Practices, Applications, and Considerations
Synthetic data generation has become an increasingly important component of methodological development in magnetic resonance spectroscopy (MRS), particularly in contexts where ground‐truth in vivo measurements are limited, noisy, or inaccessible. This review provides a structured overview of current practices in synthetic MRS data generation, with emphasis on the physical, biochemical, and computational assumptions that underpin commonly used models. A decomposition of synthetic MRS data is presented in terms of core building blocks, including basis sets, signal formation models, and metabolite parameter distributions such as concentration and relaxation properties. This framework is extended to advanced components required for realistic data synthesis, including baseline and nuisance signals, spatial encoding effects such as coil sensitivities and field inhomogeneities, k‐space sampling and gradient imperfections, as well as sample‐specific and temporal dynamics. Beyond single‐voxel modeling, synthetic data generation across multiple MRS modalities is discussed, including functional MRS, diffusion‐weighted MRS, and spectroscopic imaging, with emphasis on modality‐specific challenges such as physiological variability, motion, and confounding signal contributions. The role of synthetic data in clinical and preclinical research, acquisition optimization, software validation, and emerging artificial intelligence applications is further examined, with particular attention to domain shift and representational realism. Current practices in validation, reporting standards, and data accessibility are also reviewed, and key gaps limiting reproducibility and interoperability across studies are identified. Overall, this work provides a unified conceptual framework for synthetic MRS data generation, clarifies underlying assumptions across modeling choices, and outlines open challenges toward improving realism, standardization, and utility in both methodological and translational applications. Conducted by the MRS Synthetic Data Working Group under the Code & Data Sharing Committee of the MRS Study Group of the International Society for Magnetic Resonance in Medicine (ISMRM), this review encompasses the existing literature, supplemented by the collective experience and in‐house methodologies of its contributors, providing a community perspective that extends beyond the published record.
- 17 November 2026
- Published by Wiley in NMR in Biomedicine
- Vol. 39 (12)
Open AccessSynthetic Data in MR Spectroscopy: Current Practices, Applications, and Considerations
Synthetic data generation has become an increasingly important component of methodological development in magnetic resonance spectroscopy (MRS), particularly in contexts where ground‐truth in vivo measurements are limited, noisy, or inaccessible. This review provides a structured overview of current practices in synthetic MRS data generation, with emphasis on the physical, biochemical, and computational assumptions that underpin commonly used models. A decomposition of synthetic MRS data is presented in terms of core building blocks, including basis sets, signal formation models, and metabolite parameter distributions such as concentration and relaxation properties. This framework is extended to advanced components required for realistic data synthesis, including baseline and nuisance signals, spatial encoding effects such as coil sensitivities and field inhomogeneities, k‐space sampling and gradient imperfections, as well as sample‐specific and temporal dynamics. Beyond single‐voxel modeling, synthetic data generation across multiple MRS modalities is discussed, including functional MRS, diffusion‐weighted MRS, and spectroscopic imaging, with emphasis on modality‐specific challenges such as physiological variability, motion, and confounding signal contributions. The role of synthetic data in clinical and preclinical research, acquisition optimization, software validation, and emerging artificial intelligence applications is further examined, with particular attention to domain shift and representational realism. Current practices in validation, reporting standards, and data accessibility are also reviewed, and key gaps limiting reproducibility and interoperability across studies are identified. Overall, this work provides a unified conceptual framework for synthetic MRS data generation, clarifies underlying assumptions across modeling choices, and outlines open challenges toward improving realism, standardization, and utility in both methodological and translational applications. Conducted by the MRS Synthetic Data Working Group under the Code & Data Sharing Committee of the MRS Study Group of the International Society for Magnetic Resonance in Medicine (ISMRM), this review encompasses the existing literature, supplemented by the collective experience and in‐house methodologies of its contributors, providing a community perspective that extends beyond the published record.
- Open Access
- 14 October 2026
- Published by MDPI AG in Journal of Imaging
- Vol. 12 (10) , 510
Deep Learning-Based Multiple Sclerosis Lesion Segmentation at 7 Tesla: Comparison of Single- and Multi-Contrast MRI Inputs
Automated lesion segmentation can support rapid and reproducible quantitative assessment and long-term monitoring of multiple sclerosis. Ultra-high-field 7T MRI provides increased spatial resolution and lesion visibility but introduces challenges for automated analysis, including field inhomogeneity, susceptibility-related artifacts, and spatially varying image contrast. Although deep learning methods perform well at conventional field strengths, their application to 7T MRI remains underexplored. We evaluated nnUNet using FLAIR alone, MP2RAGE alone, and combined MP2RAGE+FLAIR. The dataset included 115 longitudinal 7T MRI scans from 57 patients with MS, with 20 scans from 11 independent patients reserved for testing. DeepMedic was included as a comparative architecture. On the test set, nnUNet achieved mean DSC values of 0.72, 0.71, and 0.66 for combined, MP2RAGE-only, and FLAIR-only inputs, respectively. No significant DSC differences were observed between nnUNet and DeepMedic. For both architectures, adding FLAIR to MP2RAGE did not show a statistically significant additional benefit in the present test cohort, whereas MP2RAGE-containing inputs significantly outperformed FLAIR-only input. The findings of the present study provide preliminary evidence suggesting that the observed voxel-level segmentation performance was primarily associated with the inclusion of MP2RAGE. However, lesion-level performance depended on lesion size, indicating further investigation of contrast-specific effects on small-lesion detection.
- 14 October 2026
- Published by MDPI AG in Journal of Imaging
- Vol. 12 (10) , 510
Open AccessDeep Learning-Based Multiple Sclerosis Lesion Segmentation at 7 Tesla: Comparison of Single- and Multi-Contrast MRI Inputs
Automated lesion segmentation can support rapid and reproducible quantitative assessment and long-term monitoring of multiple sclerosis. Ultra-high-field 7T MRI provides increased spatial resolution and lesion visibility but introduces challenges for automated analysis, including field inhomogeneity, susceptibility-related artifacts, and spatially varying image contrast. Although deep learning methods perform well at conventional field strengths, their application to 7T MRI remains underexplored. We evaluated nnUNet using FLAIR alone, MP2RAGE alone, and combined MP2RAGE+FLAIR. The dataset included 115 longitudinal 7T MRI scans from 57 patients with MS, with 20 scans from 11 independent patients reserved for testing. DeepMedic was included as a comparative architecture. On the test set, nnUNet achieved mean DSC values of 0.72, 0.71, and 0.66 for combined, MP2RAGE-only, and FLAIR-only inputs, respectively. No significant DSC differences were observed between nnUNet and DeepMedic. For both architectures, adding FLAIR to MP2RAGE did not show a statistically significant additional benefit in the present test cohort, whereas MP2RAGE-containing inputs significantly outperformed FLAIR-only input. The findings of the present study provide preliminary evidence suggesting that the observed voxel-level segmentation performance was primarily associated with the inclusion of MP2RAGE. However, lesion-level performance depended on lesion size, indicating further investigation of contrast-specific effects on small-lesion detection.
- Open Access
Glutamine and NAA dissociate in ALS across somatotopically defined motor regions using 7T MRSI
Amyotrophic lateral sclerosis (ALS) is increasingly understood as a progressive neurodegenerative disorder with distributed cortical and subcortical involvement, but in vivo metabolic mapping has been limited by the spatial coverage of single-voxel proton magnetic resonance spectroscopy (MRS). We acquired high-resolution whole-brain 7T 3D-CRT-FID-MRSI alongside motor-cortex single-voxel sLASER in five rapidly progressing people living with ALS (plALS) and seven non-neurodegenerative controls (NCs), with up to three sessions per participant. Regional metabolite ratios (N-Acetylaspartate [tNAA], glutamate [Glu], glutamine [Gln] to creatine [tCr], and Glu+Gln [Glx] to tNAA) were modelled with Bayesian hierarchical mixed-effects models, and the primary motor cortex was subdivided along its dorsoventral somatotopic axis (bulbar/face, hand/upper-limb, foot/lower-limb). At baseline, plALS showed a motor-cortex-selective tNAA/tCr deficit (motor composite −8.7%, 95% credible Interval [CrI] −16.1 to −1.1, posterior probability=0.99) accompanied by cortically diffuse glutamatergic elevation (Gln/tCr +25.6%, posterior probability=0.96; Glx/tNAA +10.4%, posterior probability=0.95). Reliable separation of the J-coupled glutamine and glutamate resonances at 7T revealed Gln/tCr as a more sensitive marker of glutamatergic dysregulation than Glu/tCr alone in this cohort. Within the somatotopic subdivision, all five plALS showed their peak Gln/tCr increase in the bulbar/face zone irrespective of clinical onset, including three lower-limb-onset patients. Annualised metabolite slope by zone correlated with the matched ALSFRS-R domain decline (Glx/tNAA r=0.82, p<0.001). Group-level longitudinal interactions were modest. Bayesian assurance simulations indicated Glx/tNAA as the most efficient candidate primary endpoint for a confirmatory cross-sectional study. These findings demonstrate that 7T whole-brain MRSI can resolve a metabolic dissociation between motor-selective neuronal compromised and somatotopically patterned glutamatergic dysregulation in ALS and provide design-ready endpoint and sample-size guidance for utility as a structural biomarker of brain function in clinical trials.
Open AccessGlutamine and NAA dissociate in ALS across somatotopically defined motor regions using 7T MRSI
Amyotrophic lateral sclerosis (ALS) is increasingly understood as a progressive neurodegenerative disorder with distributed cortical and subcortical involvement, but in vivo metabolic mapping has been limited by the spatial coverage of single-voxel proton magnetic resonance spectroscopy (MRS). We acquired high-resolution whole-brain 7T 3D-CRT-FID-MRSI alongside motor-cortex single-voxel sLASER in five rapidly progressing people living with ALS (plALS) and seven non-neurodegenerative controls (NCs), with up to three sessions per participant. Regional metabolite ratios (N-Acetylaspartate [tNAA], glutamate [Glu], glutamine [Gln] to creatine [tCr], and Glu+Gln [Glx] to tNAA) were modelled with Bayesian hierarchical mixed-effects models, and the primary motor cortex was subdivided along its dorsoventral somatotopic axis (bulbar/face, hand/upper-limb, foot/lower-limb). At baseline, plALS showed a motor-cortex-selective tNAA/tCr deficit (motor composite −8.7%, 95% credible Interval [CrI] −16.1 to −1.1, posterior probability=0.99) accompanied by cortically diffuse glutamatergic elevation (Gln/tCr +25.6%, posterior probability=0.96; Glx/tNAA +10.4%, posterior probability=0.95). Reliable separation of the J-coupled glutamine and glutamate resonances at 7T revealed Gln/tCr as a more sensitive marker of glutamatergic dysregulation than Glu/tCr alone in this cohort. Within the somatotopic subdivision, all five plALS showed their peak Gln/tCr increase in the bulbar/face zone irrespective of clinical onset, including three lower-limb-onset patients. Annualised metabolite slope by zone correlated with the matched ALSFRS-R domain decline (Glx/tNAA r=0.82, p<0.001). Group-level longitudinal interactions were modest. Bayesian assurance simulations indicated Glx/tNAA as the most efficient candidate primary endpoint for a confirmatory cross-sectional study. These findings demonstrate that 7T whole-brain MRSI can resolve a metabolic dissociation between motor-selective neuronal compromised and somatotopically patterned glutamatergic dysregulation in ALS and provide design-ready endpoint and sample-size guidance for utility as a structural biomarker of brain function in clinical trials.
- Open Access
- 21 May 2026
- Published by Elsevier in SSRN Electronic Journal
7T MRSI identifies metabolically distinct multiple sclerosis lesion phenotypes associated with disability
Multiple sclerosis lesions are heterogeneous, but their in vivo metabolic characteristics remain poorly defined. We used 7T MR spectroscopic imaging to determine whether white matter lesions can be stratified based on metabolic patterns and whether these relate to paramagnetic rim status, normal-appearing white matter metabolism, and clinical disability. In this cross-sectional study, 50 people with multiple sclerosis underwent 7T MR imaging, including 3D spectroscopic imaging. After quality control, 304 white matter lesions from 45 participants were analyzed for metabolite ratios reflecting redox balance, gliosis, and neuronal integrity. Lesions were grouped using unsupervised clustering, and associations with imaging features and clinical outcomes were assessed. Three reproducible lesion clusters emerged, showing distinct metabolic patterns consistent with inflammation, metabolic exhaustion, and partial repair. Paramagnetic rim lesions were most frequent in the inflammatory cluster and least frequent in the repair-like cluster. Repair-like lesions were associated with higher antioxidative capacity in normal-appearing white matter. In contrast, lesions characterized by inflammatory and metabolic exhaustion patterns were associated with worse cognitive and motor performance, whereas repair-like lesions were not associated with disability. These findings demonstrate that 7T spectroscopic imaging identifies metabolically distinct lesion phenotypes reflecting heterogeneous pathological processes with different clinical consequences. Metabolic characterization of lesions provides clinically relevant information beyond lesion detection and may improve monitoring of disease progression in multiple sclerosis.
- 21 May 2026
- Published by Elsevier in SSRN Electronic Journal
Open Access7T MRSI identifies metabolically distinct multiple sclerosis lesion phenotypes associated with disability
Multiple sclerosis lesions are heterogeneous, but their in vivo metabolic characteristics remain poorly defined. We used 7T MR spectroscopic imaging to determine whether white matter lesions can be stratified based on metabolic patterns and whether these relate to paramagnetic rim status, normal-appearing white matter metabolism, and clinical disability. In this cross-sectional study, 50 people with multiple sclerosis underwent 7T MR imaging, including 3D spectroscopic imaging. After quality control, 304 white matter lesions from 45 participants were analyzed for metabolite ratios reflecting redox balance, gliosis, and neuronal integrity. Lesions were grouped using unsupervised clustering, and associations with imaging features and clinical outcomes were assessed. Three reproducible lesion clusters emerged, showing distinct metabolic patterns consistent with inflammation, metabolic exhaustion, and partial repair. Paramagnetic rim lesions were most frequent in the inflammatory cluster and least frequent in the repair-like cluster. Repair-like lesions were associated with higher antioxidative capacity in normal-appearing white matter. In contrast, lesions characterized by inflammatory and metabolic exhaustion patterns were associated with worse cognitive and motor performance, whereas repair-like lesions were not associated with disability. These findings demonstrate that 7T spectroscopic imaging identifies metabolically distinct lesion phenotypes reflecting heterogeneous pathological processes with different clinical consequences. Metabolic characterization of lesions provides clinically relevant information beyond lesion detection and may improve monitoring of disease progression in multiple sclerosis.
- Open Access
- 3 March 2026
- Published by Elsevier in Medical Image Analysis
- Vol. 111, 104014
PHIVE: A physics-informed variational encoder enables rapid spectral fitting of brain metabolite mapping at 7T
- 3 March 2026
- Published by Elsevier in Medical Image Analysis
- Vol. 111, 104014
Open AccessPHIVE: A physics-informed variational encoder enables rapid spectral fitting of brain metabolite mapping at 7T
- Open Access
Metabolic modelling and time-resolved mapping of glucose oxidative metabolism in rats brain by indirect deuterium detection with 1H-FID-MRSI at 9.4T
Object: The present study exploits newly developed dynamic indirect 1H-[2H]-FID-MRSI at 9.4T, combined with a dedicated metabolic model, to enable regional and quantitative characterization of glucose oxidative metabolism flux in the rat brain with minimal metabolic assumptions, by measuring both 2H-labelled Glx turnover and pool size along a controlled 2H-Glc infusion protocol. Materials and Methods: Seven rats underwent dynamic 2D 1H-FID-MRSI during a 2-hour infusion of [6,6′-2H2] glucose. Consecutive 13-min acquisitions quantified Glx-C4 1H-signal decay, converted to 2H-Glx concentrations using baseline metabolite pool sizes. A three-pool kinetic model including 2H-label loss was fitted to regional turnover curves to estimate oxidative flux (Vgt) and pyruvate dilution (Kdil). Model performance and parameter robustness were finally assessed with Monte-Carlo simulations. Results: In vivo 2H-Glx turnover showed a saturated exponential rise (~60 min), with a labelling plateau higher in striatum (1.85 μmol/g) than hippocampus (1.55 μmol/g). Metabolic modelling provided region-specific oxidative fluxes: Vgt = 0.27 ± 0.07 μmol/g/min (hippocampus) and Vgt = 0.40 ± 0.06 μmol/g/min (striatum), with consistent Kdil across regions. Simulations confirmed a good model robustness in retrieving Vgt over a large range of experimental conditions. Discussion: This work shows the appropriateness of indirect dynamic 1H-[2H]-FID-MRSI for quantitative metabolic flux mapping of cerebral glucose oxidative metabolism. Key words: Deuterium, Magnetic Resonance Spectroscopic Imaging, Glucose oxidative metabolism, Brain, Rats, QELT
Open AccessMetabolic modelling and time-resolved mapping of glucose oxidative metabolism in rats brain by indirect deuterium detection with 1H-FID-MRSI at 9.4T
Object: The present study exploits newly developed dynamic indirect 1H-[2H]-FID-MRSI at 9.4T, combined with a dedicated metabolic model, to enable regional and quantitative characterization of glucose oxidative metabolism flux in the rat brain with minimal metabolic assumptions, by measuring both 2H-labelled Glx turnover and pool size along a controlled 2H-Glc infusion protocol. Materials and Methods: Seven rats underwent dynamic 2D 1H-FID-MRSI during a 2-hour infusion of [6,6′-2H2] glucose. Consecutive 13-min acquisitions quantified Glx-C4 1H-signal decay, converted to 2H-Glx concentrations using baseline metabolite pool sizes. A three-pool kinetic model including 2H-label loss was fitted to regional turnover curves to estimate oxidative flux (Vgt) and pyruvate dilution (Kdil). Model performance and parameter robustness were finally assessed with Monte-Carlo simulations. Results: In vivo 2H-Glx turnover showed a saturated exponential rise (~60 min), with a labelling plateau higher in striatum (1.85 μmol/g) than hippocampus (1.55 μmol/g). Metabolic modelling provided region-specific oxidative fluxes: Vgt = 0.27 ± 0.07 μmol/g/min (hippocampus) and Vgt = 0.40 ± 0.06 μmol/g/min (striatum), with consistent Kdil across regions. Simulations confirmed a good model robustness in retrieving Vgt over a large range of experimental conditions. Discussion: This work shows the appropriateness of indirect dynamic 1H-[2H]-FID-MRSI for quantitative metabolic flux mapping of cerebral glucose oxidative metabolism. Key words: Deuterium, Magnetic Resonance Spectroscopic Imaging, Glucose oxidative metabolism, Brain, Rats, QELT
- Open Access
- 2 December 2025
- Published by Elsevier in NeuroImage
- Vol. 324, 121632
A deep autoencoder for fast spectral–temporal fitting of dynamic deuterium metabolic imaging data at 7T
- 2 December 2025
- Published by Elsevier in NeuroImage
- Vol. 324, 121632
Open AccessA deep autoencoder for fast spectral–temporal fitting of dynamic deuterium metabolic imaging data at 7T
- Open Access
A Deep Autoencoder for Fast Spectral-Temporal Fitting of Dynamic Deuterium Metabolic Imaging Data at 7T
Deuterium metabolic imaging (DMI) is a non-invasive magnetic resonance spectroscopic imaging technique enabling in vivo mapping of glucose metabolism. Dynamic DMI provides time-resolved metabolite maps and allows spatially resolved fitting of metabolic models to capture metabolite concentration dynamics. However, conventional fitting tools often require long processing times for high-resolution datasets, limiting their practical utility. To address this bottleneck, we propose a deep autoencoder (DAE) for rapid spectral–temporal fitting of dynamic DMI data, supporting arbitrary parametric model constraints to describe metabolite concentration dynamics. The DAE was benchmarked against spectral-temporal fitting using FSL-MRS and LCModel. Fitting accuracy was evaluated on in vivo and synthetic whole-brain dynamic DMI data acquired at 7T using Bland-Altman metrics, Pearson correlation coefficients, structural similarity index measures, and root mean squared errors for both metabolite concentrations and model constraint parameters. The DAE achieved processing times of 0.29 ms per voxel, corresponding to an acceleration of more than three orders of magnitude compared to LCModel/FSL-MRS (0.55/0.65 s per voxel). On in vivo data, it showed excellent agreement with LCModel, and on synthetic data, it consistently outperformed both reference methods across all evaluation metrics. The proposed DAE enables accurate spectral-temporal fitting for whole-brain dynamic DMI scans within less than a second, matching or exceeding the performance of conventional state-of-the-art fitting methods. This makes it a promising tool for integration into efficient post-processing pipelines for research and clinical DMI workflows.
Open AccessA Deep Autoencoder for Fast Spectral-Temporal Fitting of Dynamic Deuterium Metabolic Imaging Data at 7T
Deuterium metabolic imaging (DMI) is a non-invasive magnetic resonance spectroscopic imaging technique enabling in vivo mapping of glucose metabolism. Dynamic DMI provides time-resolved metabolite maps and allows spatially resolved fitting of metabolic models to capture metabolite concentration dynamics. However, conventional fitting tools often require long processing times for high-resolution datasets, limiting their practical utility. To address this bottleneck, we propose a deep autoencoder (DAE) for rapid spectral–temporal fitting of dynamic DMI data, supporting arbitrary parametric model constraints to describe metabolite concentration dynamics. The DAE was benchmarked against spectral-temporal fitting using FSL-MRS and LCModel. Fitting accuracy was evaluated on in vivo and synthetic whole-brain dynamic DMI data acquired at 7T using Bland-Altman metrics, Pearson correlation coefficients, structural similarity index measures, and root mean squared errors for both metabolite concentrations and model constraint parameters. The DAE achieved processing times of 0.29 ms per voxel, corresponding to an acceleration of more than three orders of magnitude compared to LCModel/FSL-MRS (0.55/0.65 s per voxel). On in vivo data, it showed excellent agreement with LCModel, and on synthetic data, it consistently outperformed both reference methods across all evaluation metrics. The proposed DAE enables accurate spectral-temporal fitting for whole-brain dynamic DMI scans within less than a second, matching or exceeding the performance of conventional state-of-the-art fitting methods. This makes it a promising tool for integration into efficient post-processing pipelines for research and clinical DMI workflows.
- Open Access
- 31 August 2025
- Published by Wiley in NMR in Biomedicine
- Vol. 38 (10) , e70125
Comparison of Low‐Rank Denoising Methods for Dynamic Deuterium MRSI at 7 T
Dynamic deuterium (2H)‐MRSI enables mapping of metabolic fluxes in vivo, but its sensitivity is hampered by the low 2H gyromagnetic ratio and 2H‐labelled metabolite concentrations. Low‐rank denoising can enhance MRSI sensitivity by separating signal from noise. Several methods have been proposed, but the optimal approach for dynamic 2H‐MRSI remains unclear. This work compares six low‐rank denoising methods for dynamic 2H‐MRSI: four variations of spatiospectral partial separability (PS)—Global PS, Local PS, Stacked PS and SPectral and temporal INtegration for SVD (SPIN‐SVD)—as well as global–local higher order singular value decomposition (GL‐HOSVD) and tensor Marchenko–Pastur principal component analysis (tMPPCA). Performance was evaluated using realistic 2H‐MRSI brain simulations across a range of noise and B0 inhomogeneity levels, including a lesion model featuring focal metabolic alteration, and in vivo data. For simulations at the in vivo SNR and B0 level, the concentration root‐mean‐square errors (RMSE) of the fitted metabolite maps against gold standard maps were reduced by 29.3%, 24.4%, 20.2%, 33.4%, 23.5% and 21.9% following denoising with Global PS, Local PS, Stacked PS, SPIN‐SVD, GL‐HOSVD and tMPPCA, respectively. Across all simulated noise levels and all but the highest B0 inhomogeneity level, SPIN‐SVD achieved the lowest concentration RMSE and best preserved spatial distributions of 2H‐labelled water (HDO) and lactate (Lac), including lesion‐associated changes. While Global PS and Stacked PS reduced noise variance, they failed to preserve signal variations in simulated and in vivo data. SPIN‐SVD, GL‐HOSVD and tMPPCA reduced noise in vivo while maintaining spatial and temporal metabolite variations. GL‐HOSVD and tMPPCA performed similarly, with tMPPCA preferred for its computational efficiency. Both SPIN‐SVD and tMPPCA are suitable for denoising dynamic 2H‐MRSI, with SPIN‐SVD preferable for clinical applications owing to its simple implementation and superior preservation of local metabolic alterations, and tMPPCA better suited for absolute quantification of very low SNR metabolites.
- 31 August 2025
- Published by Wiley in NMR in Biomedicine
- Vol. 38 (10) , e70125
Open AccessComparison of Low‐Rank Denoising Methods for Dynamic Deuterium MRSI at 7 T
Dynamic deuterium (2H)‐MRSI enables mapping of metabolic fluxes in vivo, but its sensitivity is hampered by the low 2H gyromagnetic ratio and 2H‐labelled metabolite concentrations. Low‐rank denoising can enhance MRSI sensitivity by separating signal from noise. Several methods have been proposed, but the optimal approach for dynamic 2H‐MRSI remains unclear. This work compares six low‐rank denoising methods for dynamic 2H‐MRSI: four variations of spatiospectral partial separability (PS)—Global PS, Local PS, Stacked PS and SPectral and temporal INtegration for SVD (SPIN‐SVD)—as well as global–local higher order singular value decomposition (GL‐HOSVD) and tensor Marchenko–Pastur principal component analysis (tMPPCA). Performance was evaluated using realistic 2H‐MRSI brain simulations across a range of noise and B0 inhomogeneity levels, including a lesion model featuring focal metabolic alteration, and in vivo data. For simulations at the in vivo SNR and B0 level, the concentration root‐mean‐square errors (RMSE) of the fitted metabolite maps against gold standard maps were reduced by 29.3%, 24.4%, 20.2%, 33.4%, 23.5% and 21.9% following denoising with Global PS, Local PS, Stacked PS, SPIN‐SVD, GL‐HOSVD and tMPPCA, respectively. Across all simulated noise levels and all but the highest B0 inhomogeneity level, SPIN‐SVD achieved the lowest concentration RMSE and best preserved spatial distributions of 2H‐labelled water (HDO) and lactate (Lac), including lesion‐associated changes. While Global PS and Stacked PS reduced noise variance, they failed to preserve signal variations in simulated and in vivo data. SPIN‐SVD, GL‐HOSVD and tMPPCA reduced noise in vivo while maintaining spatial and temporal metabolite variations. GL‐HOSVD and tMPPCA performed similarly, with tMPPCA preferred for its computational efficiency. Both SPIN‐SVD and tMPPCA are suitable for denoising dynamic 2H‐MRSI, with SPIN‐SVD preferable for clinical applications owing to its simple implementation and superior preservation of local metabolic alterations, and tMPPCA better suited for absolute quantification of very low SNR metabolites.
- Open Access
- 1 February 2025
- Published by Elsevier in NeuroImage
- Vol. 309, 121045
Deep-ER: Deep Learning ECCENTRIC Reconstruction for fast high-resolution neurometabolic imaging
- 1 February 2025
- Published by Elsevier in NeuroImage
- Vol. 309, 121045
Open AccessDeep-ER: Deep Learning ECCENTRIC Reconstruction for fast high-resolution neurometabolic imaging