Showing papers by "Jamie Near published in 2022"

Comparison of seven modelling algorithms for γ‐aminobutyric acid–edited proton magnetic resonance spectroscopy

Abstract: Edited MRS sequences are widely used for studying γ‐aminobutyric acid (GABA) in the human brain. Several algorithms are available for modelling these data, deriving metabolite concentration estimates through peak fitting or a linear combination of basis spectra. The present study compares seven such algorithms, using data obtained in a large multisite study. GABA‐edited (GABA+, TE = 68 ms MEGA‐PRESS) data from 222 subjects at 20 sites were processed via a standardised pipeline, before modelling with FSL‐MRS, Gannet, AMARES, QUEST, LCModel, Osprey and Tarquin, using standardised vendor‐specific basis sets (for GE, Philips and Siemens) where appropriate. After referencing metabolite estimates (to water or creatine), systematic differences in scale were observed between datasets acquired on different vendors' hardware, presenting across algorithms. Scale differences across algorithms were also observed. Using the correlation between metabolite estimates and voxel tissue fraction as a benchmark, most algorithms were found to be similarly effective in detecting differences in GABA+. An interclass correlation across all algorithms showed single‐rater consistency for GABA+ estimates of around 0.38, indicating moderate agreement. Upon inclusion of a basis set component explicitly modelling the macromolecule signal underlying the observed 3.0 ppm GABA peaks, single‐rater consistency improved to 0.44. Correlation between discrete pairs of algorithms varied, and was concerningly weak in some cases. Our findings highlight the need for consensus on appropriate modelling parameters across different algorithms, and for detailed reporting of the parameters adopted in individual studies to ensure reproducibility and meaningful comparison of outcomes between different studies.

Comparison of seven modelling algorithms for GABA-edited 1 H-MRS.

Abstract: Edited MRS sequences are widely used for studying GABA in the human brain. Several algorithms are available for modelling these data, deriving metabolite concentration estimates through peak fitting or a linear combination of basis spectra. The present study compares seven such algorithms, using data obtained in a large multi-site study. GABA-edited (GABA+, TE = 68 ms MEGA-PRESS) data from 222 subjects at 20 sites were processed via a standardised pipeline, before modelling with FSL-MRS, Gannet, AMARES, QUEST, LCModel, Osprey and Tarquin, using standardised vendor-specific basis sets (for GE, Philips and Siemens) where appropriate. After referencing metabolite estimates (to water or creatine), systematic differences in scale were observed between datasets acquired on different vendors' hardware, presenting across algorithms. Scale differences across algorithms were also observed. Using the correlation between metabolite estimates and voxel tissue fraction as a benchmark, most algorithms were found to be similarly effective in detecting differences in GABA+. An inter-class correlation across all algorithms showed single-rater consistency for GABA+ estimates of around 0.38, indicating moderate agreement. Upon inclusion of a basis set component explicitly modelling the macromolecule signal underlying the observed 3.0 ppm GABA peaks, single-rater consistency improved to 0.44. Correlation between discrete pairs of algorithms varied, and was concerningly weak in some cases. Our findings highlight the need for consensus on appropriate modelling parameters across different algorithms, and for detailed reporting of the parameters adopted in individual studies to ensure reproducibility and meaningful comparison of outcomes between different studies.

MRSCloud: A cloud‐based MRS tool for basis set simulation

Abstract: The purpose of this study is to present a cloud‐based spectral simulation tool “MRSCloud,” which allows MRS users to simulate a vendor‐specific and sequence‐specific basis set online in a convenient and time‐efficient manner. This tool can simulate basis sets for GE, Philips, and Siemens MR scanners, including conventional acquisitions and spectral editing schemes with PRESS and semi‐LASER localization at 3 T.

StandardRat: A multi-center consensus protocol to enhance functional connectivity specificity in the rat brain

Abstract: Task-free functional connectivity in animal models provides an experimental framework to examine connectivity phenomena under controlled conditions and allows comparison with invasive or terminal procedures. To date, animal acquisitions are performed with varying protocols and analyses that hamper result comparison and integration. We introduce StandardRat, a consensus rat functional MRI acquisition protocol tested across 20 centers. To develop this protocol with optimized acquisition and processing parameters, we initially aggregated 65 functional imaging datasets acquired in rats from 46 centers. We developed a reproducible pipeline for the analysis of rat data acquired with diverse protocols and determined experimental and processing parameters associated with a more robust functional connectivity detection. We show that the standardized protocol enhances biologically plausible functional connectivity patterns, relative to pre-existing acquisitions. The protocol and processing pipeline described here are openly shared with the neuroimaging community to promote interoperability and cooperation towards tackling the most important challenges in neuroscience.

Neurochemical and cognitive changes precede structural abnormalities in the TgF344-AD rat model

Abstract: Abstract Alzheimer’s disease is a progressive neurodegenerative disorder with a decades-long pre-symptomatic phase, substantiating the need for prodromal biomarker development and early intervention. To deconstruct the processes underlying disease progression and identify potential biomarkers, we used neuroimaging techniques with high translational potential to human clinical studies in the TgF344-AD rat model which recapitulates the full spectrum of Alzheimer’s neuropathology (progressive amyloid deposition, tauopathy, frank neuronal loss, gliosis, and cognitive dysfunction). We employed longitudinal MRI and magnetic resonance spectroscopy in conjunction with behavioural testing to characterize multiple facets of disease pathology in male and female TgF344-AD rats (n = 26, 14M/12F) relative to wildtype littermates (n = 24, 12M/12F). Testing was performed at 4, 10, 16, and 18 months, covering much of the adult rat lifespan and multiple stages of disease progression. The TgF344-AD model demonstrated impaired spatial reference memory in the Barnes Maze by 4 months of age, followed by neurochemical abnormalities in the hippocampus by 10 months and major structural changes by 16 months. Specifically, TgF344-AD rats displayed increased total choline and lactate, and decreased total creatine, taurine, and N-acetylaspartate to myo-inositol ratio, dentate gyrus hypertrophy, and atrophy in the hippocampus, hypothalamus, and nucleus accumbens. Overall, these findings support the use of MRI and magnetic resonance spectroscopy for the development of non-invasive biomarkers of disease progression, clarify the timing of pathological feature presentation in this model, and contribute to the validation of the TgF344-AD rat as a highly relevant model for pre-clinical Alzheimer’s disease research.

Community‐Organized Resources for Reproducible MRS Data Analysis

Abstract: In 2021, the de facto gold-standard in vivo MRS analysis software, LCModel,1 transitioned from proprietary/paid to an open-source/free model. For 29 years, this software has fulfilled the demand for robust analysis workflows, while at the same time restricting access and dissuading continued development. We argue that this moment represents a golden opportunity for MRS developers and users to transition toward collaborative community-sourced workflows and datasets. NMR spectroscopy preceded MR medical imaging by decades and offers unique potential to noninvasively extract biochemical information. However, the clinical impact of in vivo MRS is modest compared to functional, diffusion, perfusion, or magnetization transfer imaging. One particular reason is comparably low replicability and comparability of metabolite estimates across MRS groups, which is exacerbated by heterogeneous analysis approaches and poor validation.2–5 This hardly comes as a surprise as MRS spans several highly specialized subfields (X-nuclei, spectral editing, spectroscopic imaging), all operating under low SNR conditions relative to conventional MRI. Like any inherently quantitative technique, MRS estimates depend on decisions made during preprocessing, modeling, and quantification.2,3 Historically, MRS groups tend to use local data analysis infrastructure, often matured over years of research. However, local operational continuity comes at the expense of comparability, reproducibility, and repeatability of metabolite estimates across sites, vendor ecosystems, and software environments. Expert consensus recommendations on data acquisition,6–8 processing and modeling,9 macromolecule handling,10 nomenclature,11 and reporting12 have recently been published, recognizing the need for methodological harmonization. These recommendations need to be widely adopted. But shifting established lab-specific workflows to consensus practices requires effort with little incentive (or funding) to individual labs. We propose a more efficient way of implementing consensus: pooling resources to create community-organized deliverables like software tools, databases, continued support, and educational material. This paradigm is applicable for single-voxel MRS (where substantial consensus exists) and can serve as a template for spectroscopic imaging, which is methodologically more diverse and where consensus formation is less mature. There is precedent in the MR community: the fMRI field has been able to build around collaborative open-source platforms (AFNI,13 SPM,14 FSL15) from its earliest days. This allowed the community and software developers to identify weaknesses, integrate new methods into mainstream workflows, and boost standardization. As consensus recommendations emerge, a shift in the MRS field toward community organization has already occurred. Four authors of this article served as inaugural members of the Code & Data Sharing Committee of the ISMRM MR Spectroscopy Study Group, creating community resources and efforts to harmonize spectroscopic data storage into the NIfTI-MRS format16 and a BIDS17 MRS extension proposal. Open-source end-to-end analysis software18–22 has emerged for a variety of environments (Python, MATLAB, and R), including modular consensus-based preprocessing routines and customizable linear-combination modeling algorithms. The coalescence of these initiatives occurred at an auspicious moment: the transition of the LCModel software to open source and free availability. We challenge the community to look upon this convergence of events as an opportunity for MRS developers and users alike to transition toward community-organized resource building. The ultimate endpoint is an ecosystem of interoperable analysis tools and datasets, created and maintained through community projects and connected by standardized data storage specifications. This will: • encourage consensus adoption, • facilitate development and validation of new MRS methods, • accelerate integration into mainstream usage, • offer long-term accessibility and maintenance, • reduce duplicate efforts.

MRSCloud: a Cloud-based MR Spectroscopy Tool for Basis Set Simulation

Abstract: Background Accurate quantification of in vivo proton magnetic resonance spectra involves modeling with a linear combination of known metabolite basis functions. Basis sets can be generated by numerical simulation using the quantum mechanical density-matrix formalism. Accurate simulations for a basis set require correct sequence timings, and pulse shapes and durations. Purpose To present a cloud-based spectral simulation tool ‘MRSCloud’. It allows community users of MRS to simulate a vendor- and sequence-specific basis set online in a convenient and timeefficient manner. This tool can simulate basis sets for 3 major MR scanner vendors (GE, Philips, Siemens), including conventional acquisitions and spectral editing schemes (MEGA, HERMES, HERCULES) with PRESS and semi-LASER localization. Study Type Prospective. Specimen N/A Field Strength/Sequence Simulations of 3T basis sets for conventional and spectral-editing sequences (MEGA, HERMES, HERCULES) with PRESS and sLASER localizations. Assessment Simulated metabolite basis functions generated by MRSCloud are compared to those generated by FID-A and MARSS, and a phantom-acquired basis-set from LCModel. Statistical Tests Intraclass correlation coefficients (ICC) were calculated to measure the agreement between individual metabolite basis functions generated using different packages. Statistical analysis was performed using R in RStudio. Results Simulation time for a full basis set is approximately 1 hour. ICCs between MRSCloud and FID-A were at least 0.98 and ICCs between MRSCloud and MARSS were at least 0.96. ICCs between simulated MRSCloud basis spectra and acquired LCModel basis spectra were lowest for Gln at 0.68 and highest for NAA at 0.96. Data Conclusion Substantial reductions in runtime have been achieved by implementing the 1D projection method, coherence-order filtering, and pre-calculation of propagators. High ICC values indicated that the accelerating features are running correctly and produce comparable and accurate basis sets. The generated basis set has been successfully used with LCModel.

Impact of Maternal Immune Activation Early in Pregnancy on Brain Development of Offspring: A Combined Morphological, Spectroscopic, and Behavioral Study

Abstract: Serological human birth cohort studies have identified maternal infection during pregnancy as a risk factor for development of disorders such as Autism Spectrum Disorder and schizophrenia in offspring. Similarly, in experiments using animal models, maternal immune activation (MIA) has been shown to alter neuroanatomical and behavioral development in offspring. This study employs magnetic resonance imaging (MRI) and magnetic resonance spectroscopy (MRS) in conjunction with behavioral assays to refine our understanding of the impact of MIA on neurobiological development in exposed animals. On gestational day nine, pregnant dams were injected with either polyinosinic:polycytidylic acid (POL) to induce MIA or saline (SAL) as a control. Whole-brain MRI, localized proton MRS, and behavioral tests (open field, three chambered social approach, and prepulse inhibition) were acquired at two timepoints, during adolescence (postnatal day [PND] 35) and adulthood (PND 60). Whole-brain voxel-wise volumetric analyses revealed that MIA offspring exhibited altered volume in the hippocampus and caudate putamen (CPu) between adolescence and early adulthood. MRS data were assessed at each timepoint separately; MIA offspring during early adulthood but not adolescence exhibited trending reductions in γ-aminobutyrate (GABA) (p = 0.06) and myo-inositol (Ins) (p = 0.08) compared to saline controls. However, these metabolite differences did not reach levels of significance, even before multiple comparison corrections. Open field testing revealed that during adolescence, MIA offspring displayed a more anxious phenotype than controls wherein they spent less time in the anxiogenic center zone of the open field arena (p < 0.007), but this difference normalized by adulthood. There were no significant differences in sociability preference, novelty preference, or prepulse inhibition comparing the groups. Results suggest that early gestational exposure to MIA results in subtle neuroanatomical changes in the trajectories of development, trending behavioral changes in adolescent offspring, and slight neurochemical changes in young adult offspring. Maternal infection alone may not be enough; additional genetic or environmental risk factors may be required to elicit the more typical symptoms of neuropsychiatric disorders.