Multicenter reliability of diffusion tensor imaging

Investigating brain-gut microbiota dynamics and inflammatory processes in an autistic-like rat model using MRI biomarkers during childhood and adolescence

Autism spectrum disorder (ASD) is characterized by social interaction deficits and repetitive behaviors. Recent research has linked that gut dysbiosis may contribute to ASD-like behaviors. However, the exact developmental time point at which gut microbiota alterations affect brain function and behavior in patients with ASD remains unclear. We hypothesized that ASD-related brain microstructural changes and gut dysbiosis induce metabolic dysregulation and proinflammatory responses, which collectively contribute to the social behavioral deficits observed in early childhood. We used an autistic-like rat model that was generated via prenatal valproic acid exposure. We analyzed brain microstructural changes using diffusion tensor imaging (DTI) and examined microbiota, blood, and fecal samples for inflammation biomarkers. The ASD model rats exhibited significant brain microstructural changes in the anterior cingulate cortex, hippocampus, striatum, and thalamus; reduced microbiota diversity (Prevotellaceae and Peptostreptococcaceae); and altered metabolic signatures. The shift in microbiota diversity and density observed at postnatal day (PND) 35, which is a critical developmental period, underscored the importance of early ASD interventions. We identified a unique metabolic signature in the ASD model, with elevated formate and reduced acetate and butyrate levels, indicating a dysregulation in short-chain fatty acid (SCFA) metabolism. Furthermore, increased astrocytic and microglial activation and elevated proinflammatory cytokines-interleukin-1 beta (IL-1β), interleukin-6 (IL-6), interferon-gamma (IFN-γ), and tumor necrosis factor-alpha (TNF-α)-were observed, indicating immune dysregulation. This study provided insights into the complex interplay between the brain and the gut, and indicated DTI metrics as potential imaging-based biomarkers in ASD, thus emphasizing the need for early childhood interventions.

DeepResBat: Deep residual batch harmonization accounting for covariate distribution differences

Pooling MRI data from multiple datasets requires harmonization to reduce undesired inter-site variabilities, while preserving effects of biological variables (or covariates). The popular harmonization approach ComBat uses a mixed effect regression framework that explicitly accounts for covariate distribution differences across datasets. There is also significant interest in developing harmonization approaches based on deep neural networks (DNNs), such as conditional variational autoencoder (cVAE). However, current DNN approaches do not explicitly account for covariate distribution differences across datasets. Here, we provide mathematical results, suggesting that not accounting for covariates can lead to suboptimal harmonization. We propose two DNN-based covariate-aware harmonization approaches: covariate VAE (coVAE) and DeepResBat. The coVAE approach is a natural extension of cVAE by concatenating covariates and site information with site- and covariate-invariant latent representations. DeepResBat adopts a residual framework inspired by ComBat. DeepResBat first removes the effects of covariates with nonlinear regression trees, followed by eliminating site differences with cVAE. Finally, covariate effects are added back to the harmonized residuals. Using three datasets from three continents with a total of 2787 participants and 10,085 anatomical T1 scans, we find that DeepResBat and coVAE outperformed ComBat, CovBat and cVAE in terms of removing dataset differences, while enhancing biological effects of interest. However, coVAE hallucinates spurious associations between anatomical MRI and covariates even when no association exists. Future studies proposing DNN-based harmonization approaches should be aware of this false positive pitfall. Overall, our results suggest that DeepResBat is an effective deep learning alternative to ComBat. Code for DeepResBat can be found here: https://github.com/ThomasYeoLab/CBIG/tree/master/stable_projects/harmonization/An2024_DeepResBat.

Harmonization of Structural Brain Connectivity through Distribution Matching

The increasing prevalence of multi-site diffusion-weighted magnetic resonance imaging (dMRI) studies potentially offers enhanced statistical power for investigating brain structure. However, these studies face challenges due to variations in scanner hardware and acquisition protocols. While several methods exist for dMRI data harmonization, few specifically address structural brain connectivity. We introduce a new distribution-matching approach to harmonizing structural brain connectivity across different sites and scanners. We evaluate our method using structural brain connectivity data from two distinct datasets of OASIS-3 and ADNI-2, comparing its performance to the widely used ComBat method. Our approach is meant to align the statistical properties of connectivity data from these two datasets. We examine the impact of harmonization on the correlation of brain connectivity with the Mini-Mental State Examination score and age. Our results demonstrate that our distribution-matching technique more effectively harmonizes structural brain connectivity, often producing stronger and more significant correlations compared to ComBat. Qualitative assessments illustrate the desired distributional alignment of ADNI-2 with OASIS-3, while quantitative evaluations confirm robust performance. This work contributes to the growing field of dMRI harmonization, potentially improving the reliability and comparability of structural connectivity studies that combine data from different sources in neuroscientific and clinical research.

MidRISH: Unbiased harmonization of rotationally invariant harmonics of the diffusion signal

Data harmonization is necessary for removing confounding effects in multi-site diffusion image analysis. One such harmonization method, LinearRISH, scales rotationally invariant spherical harmonic (RISH) features from one site ("target") to the second ("reference") to reduce confounding scanner effects. However, reference and target site designations are not arbitrary and resultant diffusion metrics (fractional anisotropy, mean diffusivity) are biased by this choice. In this work we propose MidRISH: rather than scaling reference RISH features to target RISH features, we project both sites to a mid-space. We validate MidRISH with the following experiments: harmonizing scanner differences from 37 matched patients free of cognitive impairment, and harmonizing acquisition and study differences on 117 matched patients free of cognitive impairment. We find that MidRISH reduces bias of reference selection while preserving harmonization efficacy of LinearRISH. Users should be cautious when performing LinearRISH harmonization. To select a reference site is to choose diffusion metric effect-size. Our proposed method eliminates the bias-inducing site selection step.

DeepResBat: deep residual batch harmonization accounting for covariate distribution differences

Pooling MRI data from multiple datasets requires harmonization to reduce undesired inter-site variabilities, while preserving effects of biological variables (or covariates). The popular harmonization approach ComBat uses a mixed effect regression framework that explicitly accounts for covariate distribution differences across datasets. There is also significant interest in developing harmonization approaches based on deep neural networks (DNNs), such as conditional variational autoencoder (cVAE). However, current DNN approaches do not explicitly account for covariate distribution differences across datasets. Here, we provide mathematical results, suggesting that not accounting for covariates can lead to suboptimal harmonization. We propose two DNN-based covariate-aware harmonization approaches: covariate VAE (coVAE) and DeepResBat. The coVAE approach is a natural extension of cVAE by concatenating covariates and site information with site- and covariate-invariant latent representations. DeepResBat adopts a residual framework inspired by ComBat. DeepResBat first removes the effects of covariates with nonlinear regression trees, followed by eliminating site differences with cVAE. Finally, covariate effects are added back to the harmonized residuals. Using three datasets from three continents with a total of 2787 participants and 10085 anatomical T1 scans, we find that DeepResBat and coVAE outperformed ComBat, CovBat and cVAE in terms of removing dataset differences, while enhancing biological effects of interest. However, coVAE hallucinates spurious associations between anatomical MRI and covariates even when no association exists. Future studies proposing DNN-based harmonization approaches should be aware of this false positive pitfall. Overall, our results suggest that DeepResBat is an effective deep learning alternative to ComBat. Code for DeepResBat can be found here: https://github.com/ThomasYeoLab/CBIG/tree/master/stable_projects/harmonization/An2024_DeepResBat.

Reduced cross-scanner variability using vendor-agnostic sequences for single-shell diffusion MRI

PURPOSE

To reduce the inter-scanner variability of diffusion MRI (dMRI) measures between scanners from different vendors by developing a vendor-neutral dMRI pulse sequence using the open-source vendor-agnostic Pulseq platform.

METHODS

We implemented a standard EPI based dMRI sequence in Pulseq. We tested it on two clinical scanners from different vendors (Siemens Prisma and GE Premier), systematically evaluating and comparing the within- and inter-scanner variability across the vendors, using both the vendor-provided and Pulseq dMRI sequences. Assessments covered both a diffusion phantom and three human subjects, using standard error (SE) and Lin's concordance correlation to measure the repeatability and reproducibility of standard DTI metrics including fractional anisotropy (FA) and mean diffusivity (MD).

RESULTS

Identical dMRI sequences were executed on both scanners using Pulseq. On the phantom, the Pulseq sequence showed more than a 2.5× reduction in SE (variability) across Siemens and GE scanners. Furthermore, Pulseq sequences exhibited markedly reduced SE in-vivo, maintaining scan-rescan repeatability while delivering lower variability in FA and MD (more than 50% reduction in cortical/subcortical regions) compared to vendor-provided sequences.

CONCLUSION

The Pulseq diffusion sequence reduces the cross-scanner variability for both phantom and in-vivo data, which will benefit multi-center neuroimaging studies and improve the reproducibility of neuroimaging studies.

Diffusion MRI harmonization via personalized template mapping

One fundamental challenge in diffusion magnetic resonance imaging (dMRI) harmonization is to disentangle the contributions of scanner-related effects from the variable brain anatomy for the observed imaging signals. Conventional harmonization methods rely on establishing an atlas space to resolve anatomical variability and generate a unified inter-site mapping function. However, this approach is limited in accounting for the misalignment of neuroanatomy that still widely persists even after registration, especially in regions close to cortical boundaries. To overcome this challenge, we propose a personalized framework in this paper to more effectively address the confounding from the misalignment of neuroanatomy in dMRI harmonization. Instead of using a common template representing site-effects for all subjects, the main novelty of our method is the adaptive computation of personalized templates for both source and target scanning sites to estimate the inter-site mapping function. We integrate our method with the rotation invariant spherical harmonics (RISH) features to achieve the harmonization of dMRI signals between sites. In our experiments, the proposed approach is applied to harmonize the dMRI data acquired from two scanning platforms: Siemens Prisma and GE MR750 from the Adolescent Brain Cognitive Development dataset and compared with a state-of-the-art method based on RISH features. Our results indicate that the proposed harmonization framework achieves superior performance not only in reducing inter-site variations due to scanner differences but also in preserving sex-related biological variability in original cohorts. Moreover, we assess the impact of harmonization on the estimation of fiber orientation distributions and show the robustness of the personalized harmonization procedure in preserving the fiber orientation of original dMRI signals.

Empirical assessment of the assumptions of ComBat with diffusion tensor imaging

Purpose

Diffusion tensor imaging (DTI) is a magnetic resonance imaging technique that provides unique information about white matter microstructure in the brain but is susceptible to confounding effects introduced by scanner or acquisition differences. ComBat is a leading approach for addressing these site biases. However, despite its frequent use for harmonization, ComBat's robustness toward site dissimilarities and overall cohort size have not yet been evaluated in terms of DTI.

Approach

As a baseline, we match N = 358 participants from two sites to create a "silver standard" that simulates a cohort for multi-site harmonization. Across sites, we harmonize mean fractional anisotropy and mean diffusivity, calculated using participant DTI data, for the regions of interest defined by the JHU EVE-Type III atlas. We bootstrap 10 iterations at 19 levels of total sample size, 10 levels of sample size imbalance between sites, and 6 levels of mean age difference between sites to quantify (i) β AGE , the linear regression coefficient of the relationship between FA and age; (ii) γ ^ s f * , the ComBat-estimated site-shift; and (iii) δ ^ s f * , the ComBat-estimated site-scaling. We characterize the reliability of ComBat by evaluating the root mean squared error in these three metrics and examine if there is a correlation between the reliability of ComBat and a violation of assumptions.

Results

ComBat remains well behaved for β AGE when N > 162 and when the mean age difference is less than 4 years. The assumptions of the ComBat model regarding the normality of residual distributions are not violated as the model becomes unstable.

Conclusion

Prior to harmonization of DTI data with ComBat, the input cohort should be examined for size and covariate distributions of each site. Direct assessment of residual distributions is less informative on stability than bootstrap analysis. We caution use ComBat of in situations that do not conform to the above thresholds.

Learning site-invariant features of connectomes to harmonize complex network measures

Multi-site diffusion MRI data is often acquired on different scanners and with distinct protocols. Differences in hardware and acquisition result in data that contains site dependent information, which confounds connectome analyses aiming to combine such multi-site data. We propose a data-driven solution that isolates site-invariant information whilst maintaining relevant features of the connectome. We construct a latent space that is uncorrelated with the imaging site and highly correlated with patient age and a connectome summary measure. Here, we focus on network modularity. The proposed model is a conditional, variational autoencoder with three additional prediction tasks: one for patient age, and two for modularity trained exclusively on data from each site. This model enables us to 1) isolate site-invariant biological features, 2) learn site context, and 3) re-inject site context and project biological features to desired site domains. We tested these hypotheses by projecting 77 connectomes from two studies and protocols (Vanderbilt Memory and Aging Project (VMAP) and Biomarkers of Cognitive Decline Among Normal Individuals (BIOCARD) to a common site. We find that the resulting dataset of modularity has statistically similar means (p-value <0.05) across sites. In addition, we fit a linear model to the joint dataset and find that positive correlations between age and modularity were preserved.

Evaluation of Mean Shift, ComBat, and CycleGAN for Harmonizing Brain Connectivity Matrices Across Sites

Connectivity matrices derived from diffusion MRI (dMRI) provide an interpretable and generalizable way of understanding the human brain connectome. However, dMRI suffers from inter-site and between-scanner variation, which impedes analysis across datasets to improve robustness and reproducibility of results. To evaluate different harmonization approaches on connectivity matrices, we compared graph measures derived from these matrices before and after applying three harmonization techniques: mean shift, ComBat, and CycleGAN. The sample comprises 168 age-matched, sex-matched normal subjects from two studies: the Vanderbilt Memory and Aging Project (VMAP) and the Biomarkers of Cognitive Decline Among Normal Individuals (BIOCARD). First, we plotted the graph measures and used coefficient of variation (CoV) and the Mann-Whitney U test to evaluate different methods' effectiveness in removing site effects on the matrices and the derived graph measures. ComBat effectively eliminated site effects for global efficiency and modularity and outperformed the other two methods. However, all methods exhibited poor performance when harmonizing average betweenness centrality. Second, we tested whether our harmonization methods preserved correlations between age and graph measures. All methods except for CycleGAN in one direction improved correlations between age and global efficiency and between age and modularity from insignificant to significant with p-values less than 0.05.

Evaluation of Mean Shift, ComBat, and CycleGAN for Harmonizing Brain Connectivity Matrices Across Sites

Connectivity matrices derived from diffusion MRI (dMRI) provide an interpretable and generalizable way of understanding the human brain connectome. However, dMRI suffers from inter-site and between-scanner variation, which impedes analysis across datasets to improve robustness and reproducibility of results. To evaluate different harmonization approaches on connectivity matrices, we compared graph measures derived from these matrices before and after applying three harmonization techniques: mean shift, ComBat, and CycleGAN. The sample comprises 168 age-matched, sex-matched normal subjects from two studies: the Vanderbilt Memory and Aging Project (VMAP) and the Biomarkers of Cognitive Decline Among Normal Individuals (BIOCARD). First, we plotted the graph measures and used coefficient of variation (CoV) and the Mann-Whitney U test to evaluate different methods' effectiveness in removing site effects on the matrices and the derived graph measures. ComBat effectively eliminated site effects for global efficiency and modularity and outperformed the other two methods. However, all methods exhibited poor performance when harmonizing average betweenness centrality. Second, we tested whether our harmonization methods preserved correlations between age and graph measures. All methods except for CycleGAN in one direction improved correlations between age and global efficiency and between age and modularity from insignificant to significant with p-values less than 0.05.

MidRISH: Unbiased harmonization of rotationally invariant harmonics of the diffusion signal

Objective

Data harmonization is necessary for removing confounding effects in multi-site diffusion image analysis. One such harmonization method, LinearRISH, scales rotationally invariant spherical harmonic (RISH) features from one site ("target") to the second ("reference") to reduce confounding scanner effects. However, reference and target site designations are not arbitrary and resultant diffusion metrics (fractional anisotropy, mean diffusivity) are biased by this choice. In this work we propose MidRISH: rather than scaling reference RISH features to target RISH features, we project both sites to a mid-space.

Methods

We validate MidRISH with the following experiments: harmonizing scanner differences from 37 matched patients free of cognitive impairment, and harmonizing acquisition and study differences on 117 matched patients free of cognitive impairment.

Conclusion

MidRISH reduces bias of reference selection while preserving harmonization efficacy of LinearRISH.

Significance

Users should be cautious when performing LinearRISH harmonization. To select a reference site is to choose diffusion metric effect-size. Our proposed method eliminates the bias-inducing site selection step.

Mixed longitudinal and cross-sectional analyses of deep gray matter and white matter using diffusion weighted images in premanifest and manifest Huntington's disease

Changes in the brain of patients with Huntington's disease (HD) begin years before clinical onset, so it remains critical to identify biomarkers to track these early changes. Metrics derived from tensor modeling of diffusion-weighted MRIs (DTI), that indicate the microscopic brain structure, can add important information to regional volumetric measurements. This study uses two large-scale longitudinal, multicenter datasets, PREDICT-HD and IMAGE-HD, to trace changes in DTI of HD participants with a broad range of CAP scores (a product of CAG repeat expansion and age), including those with pre-manifest disease (i.e., prior to clinical onset). Utilizing a fully automated data-driven approach to study the whole brain divided in regions of interest, we traced changes in DTI metrics (diffusivity and fractional anisotropy) versus CAP scores, using sigmoidal and linear regression models. We identified points of inflection in the sigmoidal regression using change-point analysis. The deep gray matter showed more evident and earlier changes in DTI metrics over CAP scores, compared to the deep white matter. In the deep white matter, these changes were more evident and occurred earlier in superior and posterior areas, compared to anterior and inferior areas. The curves of mean diffusivity vs. age of HD participants within a fixed CAP score were different from those of controls, indicating that the disease has an additional effect to age on the microscopic brain structure. These results show the regional and temporal vulnerability of the white matter and deep gray matter in HD, with potential implications for experimental therapeutics.

Inter-vendor and inter-observer reliability of diffusion tensor imaging in the musculoskeletal system: a multiscanner MR study

BACKGROUND

To evaluate the inter-observer and inter-vendor reliability of diffusion tensor imaging parameters in the musculoskeletal system.

METHODS

This prospective study included six healthy volunteers three men (mean age: 42; range: 31-52 years) and three women (mean age: 36; range: 30-44 years). Each subject was scanned using different 3 Tesla magnetic resonance scanners from three different vendors at three different sites bilaterally. First, the intra-class correlation coefficient was used to determine between-observers agreement for overall measurements and clinical sites. Next, between-group comparisons were made through the nonparametric Friedman's test. Finally, the Bland-Altman method was used to determine agreement among the three scanner measurements, comparing them two by two.

RESULTS

A total of 792 measurement were calculated. ICC reported high levels of agreement between the two observers. ICC related to MD, FA, and RD measurements ranged from 0.88 (95% CI 0.85-0.90) to 0.95 (95% CI 0.94-0.96), from 0.85 (95% CI 0.81-0.88) to 0.95 (95% CI 0.93-0.96), and from 0.89 (0.85-0.90) to 0.92 (0.90-0.94). No statistically significant inter-vendor differences were observed. The Bland-Altmann method confirmed a high correlation between parameter values.

CONCLUSION

An excellent inter-observer and inter-vendor reliability was found in our study.

DBSegment: Fast and robust segmentation of deep brain structures considering domain generalization

Segmenting deep brain structures from magnetic resonance images is important for patient diagnosis, surgical planning, and research. Most current state-of-the-art solutions follow a segmentation-by-registration approach, where subject magnetic resonance imaging (MRIs) are mapped to a template with well-defined segmentations. However, registration-based pipelines are time-consuming, thus, limiting their clinical use. This paper uses deep learning to provide a one-step, robust, and efficient deep brain segmentation solution directly in the native space. The method consists of a preprocessing step to conform all MRI images to the same orientation, followed by a convolutional neural network using the nnU-Net framework. We use a total of 14 datasets from both research and clinical collections. Of these, seven were used for training and validation and seven were retained for testing. We trained the network to segment 30 deep brain structures, as well as a brain mask, using labels generated from a registration-based approach. We evaluated the generalizability of the network by performing a leave-one-dataset-out cross-validation, and independent testing on unseen datasets. Furthermore, we assessed cross-domain transportability by evaluating the results separately on different domains. We achieved an average dice score similarity of 0.89 ± 0.04 on the test datasets when compared to the registration-based gold standard. On our test system, the computation time decreased from 43 min for a reference registration-based pipeline to 1.3 min. Our proposed method is fast, robust, and generalizes with high reliability. It can be extended to the segmentation of other brain structures. It is publicly available on GitHub, and as a pip package for convenient usage.

Comparing voxel- and feature-wise harmonization of complex graph measures from multiple sites for structural brain network investigation of aging

Complex graph theory measures of brain structural connectomes derived from diffusion weighted images (DWI) provide insight into the network structure of the brain. Further, as the number of available DWI datasets grows, so does the ability to investigate associations in these measures with major biological factors, like age. However, one key hurdle that remains is the presence of scanner effects that can arise from different DWI datasets and confound multisite analyses. Two common approaches to correct these effects are voxel-wise and feature-wise harmonization. However, it is still unclear how to best leverage them for graph-theory analysis of an aging population. Thus, there is a need to better characterize the impact of each harmonization method and their ability to preserve age related features. We investigate this by characterizing four complex graph theory measures (modularity, characteristic path length, global efficiency, and betweenness centrality) in 48 participants aged 55 to 86 from Baltimore Longitudinal Study of Aging (BLSA) and Vanderbilt Memory and Aging Project (VMAP) before and after voxel- and feature-wise harmonization with the Null Space Deep Network (NSDN) and ComBat, respectively. First, we characterize across dataset coefficients of variation (CoV) and find the combination of NSDN and ComBat causes the greatest reduction in CoV followed by ComBat alone then NSDN alone. Second, we reproduce published associations of modularity with age after correcting for other covariates with linear models. We find that harmonization with ComBat or ComBat and NSDN together improves the significance of existing age effects, reduces model residuals, and qualitatively reduces separation between datasets. These results reinforce the efficiency of statistical harmonization on the feature-level with ComBat and suggest that harmonization on the voxel-level is synergistic but may have reduced effect after running through the multiple layers of the connectomics pipeline. Thus, we conclude that feature-wise harmonization improves statistical results, but the addition of biologically informed voxel-based harmonization offers further improvement.

Goal-specific brain MRI harmonization

There is significant interest in pooling magnetic resonance image (MRI) data from multiple datasets to enable mega-analysis. Harmonization is typically performed to reduce heterogeneity when pooling MRI data across datasets. Most MRI harmonization algorithms do not explicitly consider downstream application performance during harmonization. However, the choice of downstream application might influence what might be considered as study-specific confounds. Therefore, ignoring downstream applications during harmonization might potentially limit downstream performance. Here we propose a goal-specific harmonization framework that utilizes downstream application performance to regularize the harmonization procedure. Our framework can be integrated with a wide variety of harmonization models based on deep neural networks, such as the recently proposed conditional variational autoencoder (cVAE) harmonization model. Three datasets from three different continents with a total of 2787 participants and 10,085 anatomical T1 scans were used for evaluation. We found that cVAE removed more dataset differences than the widely used ComBat model, but at the expense of removing desirable biological information as measured by downstream prediction of mini mental state examination (MMSE) scores and clinical diagnoses. On the other hand, our goal-specific cVAE (gcVAE) was able to remove as much dataset differences as cVAE, while improving downstream cross-sectional prediction of MMSE scores and clinical diagnoses.

Lifespan Volume Trajectories From Non–harmonized T1–Weighted MRI Do Not Differ After Site Correction Based on Traveling Human Phantoms

Multi–site imaging consortiums strive to increase participant numbers by pooling data across sites, but scanner related differences can bias results. This study combines data from three research MRI centers, including three different scanner models from two vendors, to examine non–harmonized T1–weighted brain imaging protocols in two cohorts. First, 23 human traveling phantoms were scanned twice each at all three sites (six scans per person; 138 scans total) to quantify within–participant variability of brain volumes (total brain, white matter, gray matter, lateral ventricles, thalamus, caudate, putamen and globus pallidus), and to calculate site–specific correction factors for each structure. Sample size calculations were used to determine the number of traveling phantoms needed to achieve effect sizes for observed differences to help guide future studies. Next, cross–sectional lifespan volume trajectories were examined in 856 healthy participants (5—91 years of age) scanned at these sites. Cross–sectional trajectories of volume versus age for each structure were then compared before and after application of traveling phantom based site–specific correction factors, as well as correction using the open–source method ComBat. Although small systematic differences between sites were observed in the traveling phantom analysis, correction for site using either method had little impact on the lifespan trajectories. Only white matter had small but significant differences in the intercept parameter after ComBat correction (but not traveling phantom based correction), while no other fits differed. This suggests that age–related changes over the lifespan outweigh systematic differences between scanners for volumetric analysis. This work will help guide pooling of multisite datasets as well as meta–analyses of data from non–harmonized protocols.

Effect of head motion-induced artefacts on the reliability of deep learning-based whole-brain segmentation

Due to their robustness and speed, recently developed deep learning-based methods have the potential to provide a faster and hence more scalable alternative to more conventional neuroimaging analysis pipelines in terms of whole-brain segmentation based on magnetic resonance (MR) images. These methods were also shown to have higher test-retest reliability, raising the possibility that they could also exhibit superior head motion tolerance. We investigated this by comparing the effect of head motion-induced artifacts in structural MR images on the consistency of segmentation performed by FreeSurfer and recently developed deep learning-based methods to a similar extent. We used state-of-the art neural network models (FastSurferCNN and Kwyk) and developed a new whole-brain segmentation pipeline (ReSeg) to examine whether reliability depends on choice of deep learning method. Structural MRI scans were collected from 110 participants under rest and active head motion and were evaluated for image quality by radiologists. Compared to FreeSurfer, deep learning-based methods provided more consistent segmentations across different levels of image quality, suggesting that they also have the advantage of providing more reliable whole-brain segmentations of MR images corrupted by motion-induced artifacts, and provide evidence for their practical applicability in the study of brain structural alterations in health and disease.

Multi-vendor and multisite evaluation of cerebrovascular reactivity mapping using hypercapnia challenge

Cerebrovascular reactivity (CVR), which measures the ability of cerebral blood vessels to dilate or constrict in response to vasoactive stimuli such as CO2 inhalation, is an important index of the brain's vascular health. Quantification of CVR using BOLD MRI with hypercapnia challenge has shown great promises in research and clinical studies. However, in order for it to be used as a potential imaging biomarker in large-scale and multi-site studies, the reliability of CO2-CVR quantification across different MRI acquisition platforms and researchers/raters must be examined. The goal of this report from the MarkVCID small vessel disease biomarkers consortium is to evaluate the reliability of CO2-CVR quantification in three studies. First, the inter-rater reliability of CO2-CVR data processing was evaluated by having raters from 5 MarkVCID sites process the same 30 CVR datasets using a cloud-based CVR data processing pipeline. Second, the inter-scanner reproducibility of CO2-CVR quantification was assessed in 10 young subjects across two scanners of different vendors. Third, test-retest repeatability was evaluated in 20 elderly subjects from 4 sites with a scan interval of less than 2 weeks. In all studies, the CO2 CVR measurements were performed using the fixed inspiration method, where the subjects wore a nose clip and a mouthpiece and breathed room air and 5% CO2 air contained in a Douglas bag alternatively through their mouth. The results showed that the inter-rater CoV of CVR processing was 0.08 ± 0.08% for whole-brain CVR values and ranged from 0.16% to 0.88% in major brain regions, with ICC of absolute agreement above 0.9959 for all brain regions. Inter-scanner CoV was found to be 6.90 ± 5.08% for whole-brain CVR values, and ranged from 4.69% to 12.71% in major brain regions, which are comparable to intra-session CoVs obtained from the same scanners on the same day. ICC of consistency between the two scanners was 0.8498 for whole-brain CVR and ranged from 0.8052 to 0.9185 across major brain regions. In the test-retest evaluation, test-retest CoV across different days was found to be 18.29 ± 17.12% for whole-brain CVR values, and ranged from 16.58% to 19.52% in major brain regions, with ICC of absolute agreement ranged from 0.6480 to 0.7785. These results demonstrated good inter-rater, inter-scanner, and test-retest reliability in healthy volunteers, and suggested that CO2-CVR has suitable instrumental properties for use as an imaging biomarker of cerebrovascular function in multi-site and longitudinal observational studies and clinical trials.

MASiVar: Multisite, multiscanner, and multisubject acquisitions for studying variability in diffusion weighted MRI

PURPOSE

Diffusion-weighted imaging allows investigators to identify structural, microstructural, and connectivity-based differences between subjects, but variability due to session and scanner biases is a challenge.

METHODS

To investigate DWI variability, we present MASiVar, a multisite data set consisting of 319 diffusion scans acquired at 3 T from b = 1000 to 3000 s/mm2 across 14 healthy adults, 83 healthy children (5 to 8 years), three sites, and four scanners as a publicly available, preprocessed, and de-identified data set. With the adult data, we demonstrate the capacity of MASiVar to simultaneously quantify the intrasession, intersession, interscanner, and intersubject variability of four common DWI processing approaches: (1) a tensor signal representation, (2) a multi-compartment neurite orientation dispersion and density model, (3) white-matter bundle segmentation, and (4) structural connectomics. Respectively, we evaluate region-wise fractional anisotropy, mean diffusivity, and principal eigenvector; region-wise CSF volume fraction, intracellular volume fraction, and orientation dispersion index; bundle-wise shape, volume, fractional anisotropy, and length; and whole connectome correlation and maximized modularity, global efficiency, and characteristic path length.

RESULTS

We plot the variability in these measures at each level and find that it consistently increases with intrasession to intersession to interscanner to intersubject effects across all processing approaches and that sometimes interscanner variability can approach intersubject variability.

CONCLUSIONS

This study demonstrates the potential of MASiVar to more globally investigate DWI variability across multiple levels and processing approaches simultaneously and suggests harmonization between scanners for multisite analyses should be considered before inference of group differences on subjects.

Fiber tractography bundle segmentation depends on scanner effects, vendor effects, acquisition resolution, diffusion sampling scheme, diffusion sensitization, and bundle segmentation workflow

When investigating connectivity and microstructure of white matter pathways of the brain using diffusion tractography bundle segmentation, it is important to understand potential confounds and sources of variation in the process. While cross-scanner and cross-protocol effects on diffusion microstructure measures are well described (in particular fractional anisotropy and mean diffusivity), it is unknown how potential sources of variation effect bundle segmentation results, which features of the bundle are most affected, where variability occurs, nor how these sources of variation depend upon the method used to reconstruct and segment bundles. In this study, we investigate six potential sources of variation, or confounds, for bundle segmentation: variation (1) across scan repeats, (2) across scanners, (3) across vendors (4) across acquisition resolution, (5) across diffusion schemes, and (6) across diffusion sensitization. We employ four different bundle segmentation workflows on two benchmark multi-subject cross-scanner and cross-protocol databases, and investigate reproducibility and biases in volume overlap, shape geometry features of fiber pathways, and microstructure features within the pathways. We find that the effects of acquisition protocol, in particular acquisition resolution, result in the lowest reproducibility of tractography and largest variation of features, followed by vendor-effects, scanner-effects, and finally diffusion scheme and b-value effects which had similar reproducibility as scan-rescan variation. However, confounds varied both across pathways and across segmentation workflows, with some bundle segmentation workflows more (or less) robust to sources of variation. Despite variability, bundle dissection is consistently able to recover the same location of pathways in the deep white matter, with variation at the gray matter/ white matter interface. Next, we show that differences due to the choice of bundle segmentation workflows are larger than any other studied confound, with low-to-moderate overlap of the same intended pathway when segmented using different methods. Finally, quantifying microstructure features within a pathway, we show that tractography adds variability over-and-above that which exists due to noise, scanner effects, and acquisition effects. Overall, these confounds need to be considered when harmonizing diffusion datasets, interpreting or combining data across sites, and when attempting to understand the successes and limitations of different methodologies in the design and development of new tractography or bundle segmentation methods.

Quantitative magnetic resonance imaging towards clinical application in multiple sclerosis

Quantitative MRI provides biophysical measures of the microstructural integrity of the CNS, which can be compared across CNS regions, patients, and centres. In patients with multiple sclerosis, quantitative MRI techniques such as relaxometry, myelin imaging, magnetization transfer, diffusion MRI, quantitative susceptibility mapping, and perfusion MRI, complement conventional MRI techniques by providing insight into disease mechanisms. These include: (i) presence and extent of diffuse damage in CNS tissue outside lesions (normal-appearing tissue); (ii) heterogeneity of damage and repair in focal lesions; and (iii) specific damage to CNS tissue components. This review summarizes recent technical advances in quantitative MRI, existing pathological validation of quantitative MRI techniques, and emerging applications of quantitative MRI to patients with multiple sclerosis in both research and clinical settings. The current level of clinical maturity of each quantitative MRI technique, especially regarding its integration into clinical routine, is discussed. We aim to provide a better understanding of how quantitative MRI may help clinical practice by improving stratification of patients with multiple sclerosis, and assessment of disease progression, and evaluation of treatment response.

Harmonization of diffusion MRI data sets with adaptive dictionary learning

Diffusion magnetic resonance imaging can indirectly infer the microstructure of tissues and provide metrics subject to normal variability in a population. Potentially abnormal values may yield essential information to support analysis of controls and patients cohorts, but subtle confounds could be mistaken for purely biologically driven variations amongst subjects. In this work, we propose a new harmonization algorithm based on adaptive dictionary learning to mitigate the unwanted variability caused by different scanner hardware while preserving the natural biological variability of the data. Our harmonization algorithm does not require paired training data sets, nor spatial registration or matching spatial resolution. Overcomplete dictionaries are learned iteratively from all data sets at the same time with an adaptive regularization criterion, removing variability attributable to the scanners in the process. The obtained mapping is applied directly in the native space of each subject toward a scanner-space. The method is evaluated with a public database which consists of two different protocols acquired on three different scanners. Results show that the effect size of the four studied diffusion metrics is preserved while removing variability attributable to the scanner. Experiments with alterations using a free water compartment, which is not simulated in the training data, shows that the modifications applied to the diffusion weighted images are preserved in the diffusion metrics after harmonization, while still reducing global variability at the same time. The algorithm could help multicenter studies pooling their data by removing scanner specific confounds, and increase statistical power in the process.

FastSurfer - A fast and accurate deep learning based neuroimaging pipeline

Traditional neuroimage analysis pipelines involve computationally intensive, time-consuming optimization steps, and thus, do not scale well to large cohort studies with thousands or tens of thousands of individuals. In this work we propose a fast and accurate deep learning based neuroimaging pipeline for the automated processing of structural human brain MRI scans, replicating FreeSurfer's anatomical segmentation including surface reconstruction and cortical parcellation. To this end, we introduce an advanced deep learning architecture capable of whole-brain segmentation into 95 classes. The network architecture incorporates local and global competition via competitive dense blocks and competitive skip pathways, as well as multi-slice information aggregation that specifically tailor network performance towards accurate segmentation of both cortical and subcortical structures. Further, we perform fast cortical surface reconstruction and thickness analysis by introducing a spectral spherical embedding and by directly mapping the cortical labels from the image to the surface. This approach provides a full FreeSurfer alternative for volumetric analysis (in under 1 ​min) and surface-based thickness analysis (within only around 1 ​h runtime). For sustainability of this approach we perform extensive validation: we assert high segmentation accuracy on several unseen datasets, measure generalizability and demonstrate increased test-retest reliability, and high sensitivity to group differences in dementia.

Serial MRI studies over 12 months using manual and atlas-based region of interest in patients with amyotrophic lateral sclerosis

BACKGROUND

Amyotrophic lateral sclerosis (ALS) is a neurodegenerative disease characterized by loss of upper and lower motor neurons. There is a need for an imaging biomarker to track disease progression. Previously, magnetic resonance imaging (MRI) has shown loss of grey and white matter in the brain of patients with ALS compared to controls. We performed serial diffusion tractography imaging (DTI) study of patients with ALS looking for changes over time.

METHODS

On all subjects (n = 15), we performed three MRI studies at 6 month intervals. DTI changes were assessed with tract-based spatial statistics (TBSS) and region of interest (ROI) studies. Cortic-spinal tract (CST) was selected for our ROI at the upper level; the posterior limb of internal capsule (PLIC), and a lower level in the pons.

RESULTS

There was no significant change in DTI measures over 12 months of observation. Better correlation of manual and atlas-based ROI methods was found in the posterior limb of the internal capsule than the pons.

CONCLUSION

While previous DTI studies showed significant differences between ALS subjects and controls, within individual subjects there is little evidence of progression over 12 months. This suggests that DTI is not a suitable biomarker to assess disease progression in ALS.

Distortion correction of diffusion weighted MRI without reverse phase-encoding scans or field-maps

Diffusion magnetic resonance images may suffer from geometric distortions due to susceptibility induced off resonance fields, which cause geometric mismatch with anatomical images and ultimately affect subsequent quantification of microstructural or connectivity indices. State-of-the art diffusion distortion correction methods typically require data acquired with reverse phase encoding directions, resulting in varying magnitudes and orientations of distortion, which allow estimation of an undistorted volume. Alternatively, additional field maps acquisitions can be used along with sequence information to determine warping fields. However, not all imaging protocols include these additional scans and cannot take advantage of state-of-the art distortion correction. To avoid additional acquisitions, structural MRI (undistorted scans) can be used as registration targets for intensity driven correction. In this study, we aim to (1) enable susceptibility distortion correction with historical and/or limited diffusion datasets that do not include specific sequences for distortion correction and (2) avoid the computationally intensive registration procedure typically required for distortion correction using structural scans. To achieve these aims, we use deep learning (3D U-nets) to synthesize an undistorted b0 image that matches geometry of structural T1w images and intensity contrasts from diffusion images. Importantly, the training dataset is heterogenous, consisting of varying acquisitions of both structural and diffusion. We apply our approach to a withheld test set and show that distortions are successfully corrected after processing. We quantitatively evaluate the proposed distortion correction and intensity-based registration against state-of-the-art distortion correction (FSL topup). The results illustrate that the proposed pipeline results in b0 images that are geometrically similar to non-distorted structural images, and more closely match state-of-the-art correction with additional acquisitions. In addition, we show generalizability of the proposed approach to datasets that were not in the original training / validation / testing datasets. These datasets included varying populations, contrasts, resolutions, and magnitudes and orientations of distortion and show efficacious distortion correction. The method is available as a Singularity container, source code, and an executable trained model to facilitate evaluation.

Scanner invariant representations for diffusion MRI harmonization

Purpose

In the present work, we describe the correction of diffusion‐weighted MRI for site and scanner biases using a novel method based on invariant representation.

Theory and Methods

Pooled imaging data from multiple sources are subject to variation between the sources. Correcting for these biases has become very important as imaging studies increase in size and multi‐site cases become more common. We propose learning an intermediate representation invariant to site/protocol variables, a technique adapted from information theory‐based algorithmic fairness; by leveraging the data processing inequality, such a representation can then be used to create an image reconstruction that is uninformative of its original source, yet still faithful to underlying structures. To implement this, we use a deep learning method based on variational auto‐encoders (VAE) to construct scanner invariant encodings of the imaging data.

Results

To evaluate our method, we use training data from the 2018 MICCAI Computational Diffusion MRI (CDMRI) Challenge Harmonization dataset. Our proposed method shows improvements on independent test data relative to a recently published baseline method on each subtask, mapping data from three different scanning contexts to and from one separate target scanning context.

Conclusions

As imaging studies continue to grow, the use of pooled multi‐site imaging will similarly increase. Invariant representation presents a strong candidate for the harmonization of these data.

Image acquisition and quality assurance in the Boston Adolescent Neuroimaging of Depression and Anxiety study

The Connectomes Related to Human Diseases (CRHD) initiative was developed with the Human Connectome Project (HCP) to provide high-resolution, open-access, multi-modal MRI data to better understand the neural correlates of human disease. Here, we present an introduction to a CRHD project, the Boston Adolescent Neuroimaging of Depression and Anxiety (BANDA) study, which is collecting multimodal neuroimaging, clinical, and neuropsychological data from 225 adolescents (ages 14-17), 150 of whom are expected to have a diagnosis of depression and/or anxiety. Our transdiagnostic recruitment approach samples the full spectrum of depressed/anxious symptoms and their comorbidity, consistent with NIMH Research Domain Criteria (RDoC). We focused on an age range that is critical for brain development and for the onset of mental illness. This project sought to harmonize imaging sequences, hardware, and functional tasks with other HCP studies, although some changes were made to canonical HCP methods to accommodate our study population and questions. We present a thorough overview of our imaging sequences, hardware, and scanning protocol. We detail similarities and differences between this study and other HCP studies. We evaluate structural-, diffusion-, and functional-image-quality measures that may be influenced by clinical factors (e.g., disorder, symptomatology). Signal-to-noise and motion estimates from the first 140 adolescents suggest minimal influence of clinical factors on image quality. We anticipate enrollment of an additional 85 participants, most of whom are expected to have a diagnosis of anxiety and/or depression. Clinical and neuropsychological data from the first 140 participants are currently freely available through the National Institute of Mental Health Data Archive (NDA).

Test-retest reliability and sample size estimates after MRI scanner relocation

OBJECTIVE

Many factors can contribute to the reliability and robustness of MRI-derived metrics. In this study, we assessed the reliability and reproducibility of three MRI modalities after an MRI scanner was relocated to a new hospital facility.

METHODS

Twenty healthy volunteers (12 females, mean age (standard deviation) ​= ​41 (11) years, age range [25-66]) completed three MRI sessions. The first session (S1) was one week prior to the 3T GE HDxt scanner relocation. The second (S2) occurred nine weeks after S1 and at the new location; a third session (S3) was acquired 4 weeks after S2. At each session, we acquired structural T1-weighted, pseudo-continuous arterial spin labelled, and diffusion tensor imaging sequences. We used longitudinal processing streams to create 12 summary MRI metrics, including total gray matter (GM), cortical GM, subcortical GM, white matter (WM), and lateral ventricle volume; mean cortical thickness; total surface area; average gray matter perfusion, and average diffusion tensor metrics along principal white matter pathways. We compared mean MRI values and variance at the old scanner location to multiple sessions at the new location using Bayesian multi-level regression models. K-fold cross validation allowed identification of important predictors. Whole-brain analyses were used to investigate any regional differences. Furthermore, we calculated within-subject coefficient of variation (wsCV), intraclass correlation coefficient (ICC), and dice similarity index (SI) of cortical segmentations across scanner relocation and within-site. Additionally, we estimated sample sizes required to robustly detect a 4% difference between two groups across MRI metrics.

RESULTS

All global MRI metrics exhibited little mean difference and small variability (bar cortical gray matter perfusion) both across scanner relocation and within-site repeat. T1- and DTI-derived tissue metrics showed ​< ​|0.3|% mean difference and <1.2% variance across scanner location and <|0.4|% mean difference and <0.8% variance within the new location, with between-site intraclass correlation coefficient (ICC) ​> ​0.80 and within-subject coefficient of variation (wsCV) ​< ​1.4%. Mean cortical gray matter perfusion had the highest between-session variability (6.7% [0.3, 16.7], estimate [95% uncertainty interval]), and hence the smallest ICC (0.71 [0.44,0.92]) and largest wsCV (13.4% [5.4, 18.1]). No global metric exhibited evidence of a meaningful mean difference between scanner locations. However, surface area showed evidence of a mean difference within-site repeat (between S2 and S3). Whole-brain analyses revealed no significant areas of difference between scanner relocation or within-site. For all metrics, we found no support for a systematic difference in variance across relocation sites compared to within-site test-retest reliability. Necessary sample sizes to detect a 4% difference between two independent groups varied from a maximum of n ​= ​362 per group (cortical gray matter perfusion), to total gray matter volume (n ​= ​114), average fractional anisotropy (n ​= ​23), total gray matter volume normalized by intracranial volume (n ​= ​19), and axial diffusivity (n ​= ​3 per group).

CONCLUSION

Cortical gray matter perfusion was the most variable metric investigated (necessitating large sample sizes to identify group differences), with other metrics showing substantially less variability. Scanner relocation appeared to have a negligible effect on variability of the global MRI metrics tested. This manuscript reports within-site test-retest variability to act as a tool for calculating sample size in future investigations. Our results suggest that when all other parameters are held constant (e.g., sequence parameters and MRI processing), the effect of scanner relocation is indistinguishable from within-site variability, but may need to be considered depending on the question being investigated.

White matter development in infants at risk for schizophrenia

BACKGROUND

Schizophrenia is considered a neurodevelopmental disorder with a pathophysiology that likely begins long before the onset of clinical symptoms. White matter abnormalities have been observed in schizophrenia and we hypothesized that the first 2 years of life is a period in which white matter abnormalities associated with schizophrenia risk may emerge.

METHODS

38 infants at high risk for schizophrenia and 202 healthy controls underwent diffusion tensor MRIs after birth and at 1 and 2 years of age. Quantitative tractography was used to determine diffusion properties (fractional anisotropy (FA), axial diffusivity (AD), and radial diffusivity (RD)) of 18 white matter tracts and a general linear model was used to analyze group differences at each age.

RESULTS

Adjusting gestational age at birth, postnatal age at MRI, gender, MRI scanner type, and maternal education, neonates at high risk had significantly lower FA (p = 0.02) and AD (p = 0.03) in the superior segment of the left cingulate, and higher RD in the hippocampal segment of the left cingulate (p = 0.04). High risk one year olds had significantly lower FA (p < 0.01) and AD (p = 0.02) in the hippocampal segment of the left cingulate. High risk two year olds had significantly lower FA in the left prefrontal cortico-thalamic tract (p = 0.04) and higher RD in the right uncinate fasciculus (p = 0.04). None of the tract differences remained significant after correction for multiple comparisons.

CONCLUSIONS

There is evidence of abnormal white matter development in young children at risk for schizophrenia, especially in the hippocampal segment of left cingulum. These results support the neurodevelopmental theory of schizophrenia and indicate that impaired white matter may be present in early childhood.

Reproducibility of Structural and Diffusion Tensor Imaging in the TACERN Multi-Center Study

BACKGROUND

Multi-site MRI studies are often necessary for recruiting sufficiently sized samples when studying rare conditions. However, they require pooling data from multiple scanners into a single data set, and therefore it is critical to evaluate the variability of quantitative MRI measures within and across scanners used in multi-site studies. The aim of this study was to evaluate the reproducibility of structural and diffusion weighted (DW) MRI measurements acquired on seven scanners at five medical centers as part of the Tuberous Sclerosis Complex Autism Center of Excellence Research Network (TACERN) multisite study.

METHODS

The American College of Radiology (ACR) phantom was imaged monthly to measure reproducibility of signal intensity and uniformity within and across seven 3T scanners from General Electric, Philips, and Siemens vendors. One healthy adult male volunteer was imaged repeatedly on all seven scanners under the TACERN structural and DW protocol (5 b = 0 s/mm2 and 30 b = 1000 s/mm2) over a period of 5 years (age 22-27 years). Reproducibility of inter- and intra-scanner brain segmentation volumes and diffusion tensor imaging metrics fractional anisotropy (FA) and mean diffusivity (MD) within white matter regions was quantified with coefficient of variation.

RESULTS

The American College of Radiology Phantom signal intensity and uniformity were similar across scanners and changed little over time, with a mean intra-scanner coefficient of variation of 3.6 and 1.8%, respectively. The mean inter- and intra-scanner coefficients of variation of brain structure volumes derived from T1-weighted (T1w) images of the human phantom were 3.3 and 1.1%, respectively. The mean inter- and intra-scanner coefficients of variation of FA in white matter regions were 4.5 and 2.5%, while the mean inter- and intra-scanner coefficients of variation of MD in white matter regions were 5.4 and 1.5%.

CONCLUSION

Our results suggest that volumetric and diffusion tensor imaging (DTI) measurements are highly reproducible between and within scanners and provide typical variation amplitudes that can be used as references to interpret future findings in the TACERN network.

Multi-Site Harmonization of Diffusion MRI Data via Method of Moments

Diffusion MRI is a powerful tool for non-invasive probing of brain tissue microstructure. Recent multi-center efforts in the acquisition and analysis of diffusion MRI data significantly increase sample sizes and hence improve sensitivity and reliability in detecting subtle changes associated with development, aging, and diseases. However, discrepancies resulting from different scanner vendors, acquisition protocols, and image reconstruction algorithms can cause data incompatibility across imaging centers. In this paper, we introduce a model-free method that is based on the method of moments for the direct harmonization of diffusion MRI data to reduce site-specific variations. Our method directly harmonizes diffusion-attenuated signal without the need to fit any diffusion model. Moreover, our method allows the explicit definition of well-behaved mapping functions with properties such as invertibility, smoothness, and injectivity. We show that our method is effective in lowering the variations of diffusion scalars of traveling human phantoms scanned at different sites from 1%-3% to less than 0.9% for fractional anisotropy (FA) and mean diffusivity and from 1%-2.5% to 0.3%-1.2% for generalized FA. We also demonstrate its ability in preserving individual differences and in increasing across-site consistency in tractography and white matter connectivity.

Test-retest reliability of Diffusion Tensor Imaging metrics in neonates

Diffusion tensor imaging (DTI) has been widely used in children and adults to study the microstructural features of the brain. Its use in neonate brains has been limited. Neonate brains are almost completely unmyelinated, and this together with the tendency for babies to move during a scanning session may affect the reliability of the measurements. Here we divided a 96 direction acquisition into three segments, and analysed the intra scan test-retest reliability for pairs of segments. Each segment was subjected to a rigorous quality control, and from the surviving data we chose 25 diffusion encoding directions from each segment, and assessed the pairwise reliability of the most common DTI metrics. This pairwise reliability was assessed for data from 86 infants. We used tract-based spatial statistics (TBSS), voxelwise and ROI analysis schemes, to see potential differential effects of analysis strategy and post processing on the obtained DTI metrics. We found that intra class correlation coefficient (ICC) values were generally high (ICC > 0.80). Residual motion in the data, after quality control, was not found to associate with the diffusion metrics. The results indicate that DTI metrics from neonate data can be reliable, even at relatively low angular resolution that are common for neonate scans. The results lend confidence to the use of neonate DTI data in cross sectional and longitudinal analyses in brain white matter skeleton. Future studies should assess the reliability of fiber tracking techniques in neonate data.

Test-Retest Reliability of Diffusion Measures Extracted Along White Matter Language Fiber Bundles Using HARDI-Based Tractography

High angular resolution diffusion imaging (HARDI)-based tractography has been increasingly used in longitudinal studies on white matter macro- and micro-structural changes in the language network during language acquisition and in language impairments. However, test-retest reliability measurements are essential to ascertain that the longitudinal variations observed are not related to data processing. The aims of this study were to determine the reproducibility of the reconstruction of major white matter fiber bundles of the language network using anatomically constrained probabilistic tractography with constrained spherical deconvolution based on HARDI data, as well as to assess the test-retest reliability of diffusion measures extracted along them. Eighteen right-handed participants were scanned twice, one week apart. The arcuate, inferior longitudinal, inferior fronto-occipital, and uncinate fasciculi were reconstructed in the left and right hemispheres and the following diffusion measures were extracted along each tract: fractional anisotropy, mean, axial, and radial diffusivity, number of fiber orientations, mean length of streamlines, and volume. All fiber bundles showed good morphological overlap between the two scanning timepoints and the test-retest reliability of all diffusion measures in most fiber bundles was good to excellent. We thus propose a fairly simple, but robust, HARDI-based tractography pipeline reliable for the longitudinal study of white matter language fiber bundles, which increases its potential applicability to research on the neurobiological mechanisms supporting language.

Concurrent Cross-Sectional and Longitudinal Analyses of Multivariate White Matter Profiles and Clinical Functioning in Pre-Diagnosis Huntington Disease

BACKGROUND

Gray matter (GM) atrophy in the striatum and across the brain is a consistently reported feature of the Huntington Disease (HD) prodrome. More recently, widespread prodromal white matter (WM) degradation has also been detected. However, longitudinal WM studies are limited and conflicting, and most analyses comparing WM and clinical functioning have also been cross-sectional.

OBJECTIVE

We simultaneously assessed changes in WM and cognitive and motor functioning at various prodromal HD stages.

METHODS

Data from 1,336 (1,047 prodromal, 289 control) PREDICT-HD participants were analyzed (3,700 sessions). MRI images were used to create GM, WM, and cerebrospinal fluid probability maps. Using source-based morphometry, independent component analysis was applied to WM probability maps to extract covarying spatial patterns and their subject profiles. WM profiles were analyzed in two sets of linear mixed model (LMM) analyses: one to compare WM profiles across groups cross-sectionally and longitudinally, and one to concurrently compare WM profiles and clinical variables cross-sectionally and longitudinally within each group.

RESULTS

Findings illustrate widespread prodromal changes in GM-adjacent-WM, with premotor, supplementary motor, middle frontal and striatal changes early in the prodrome that subsequently extend sub-gyrally with progression. Motor functioning agreed most with WM until the near-onset prodromal stage, when Stroop interference was the best WM indicator. Across groups, Trail-Making Test part A outperformed other cognitive variables in its similarity to WM, particularly cross-sectionally.

CONCLUSIONS

Results suggest that distinct regions coincide with cognitive compared to motor functioning. Furthermore, at different prodromal stages, distinct regions appear to align best with clinical functioning. Thus, the informativeness of clinical measures may vary according to the type of data available (cross-sectional or longitudinal) as well as age and CAG-number.

Brain imaging in comatose survivors of cardiac arrest: Pathophysiological correlates and prognostic properties

INTRODUCTION

Hypoxic-ischemic brain injury is the main cause of death and disability of comatose patients after cardiac arrest. Early and reliable prognostication is challenging. Common prognostic tools include clinical neurological examination and electrophysiological measures. Brain imaging is well established for diagnosis of focal cerebral ischemia but has so far not found worldwide application in this patient group.

OBJECTIVE

To review the value of Computed Tomography (CT), Magnetic Resonance Imaging (MRI), and Positron Emission Tomography (PET) for early prediction of neurological outcome of comatose survivors of cardiac arrest.

METHODS

A literature search was performed to identify publications on CT, MRI or PET in comatose patients after cardiac arrest.

RESULTS

We included evidence from 51 articles, 21 on CT, 27 on MRI, 1 on CT and MRI, and 2 on PET imaging. Studies varied regarding timing of measurements, choice of determinants, and cut-off values predicting poor outcome. Most studies were small (n = 6-398) and retrospective (60%). In general, cytotoxic oedema, defined by a grey-white matter ratio <1.10, derived from CT, or MRI-diffusion weighted imaging <650 × 10^-6 mm²/s in >10% of the brain could differentiate between patients with favourable and unfavourable outcomes on a group level within 1-3 days after cardiac arrest. Advanced imaging techniques such as functional MRI or diffusion tensor imaging show promising results, but need further evaluation.

CONCLUSION

CT derived grey-white matter ratio and MRI based measures of diffusivity and connectivity hold promise to improve outcome prediction after cardiac arrest. Prospective validation studies in a multivariable approach are needed to determine the additional value for the individual patient.

Multi-site harmonization of diffusion MRI data in a registration framework

Diffusion MRI (dMRI) data acquired on different scanners varies significantly in its content throughout the brain even if the acquisition parameters are nearly identical. Thus, proper harmonization of such data sets is necessary to increase the sample size and thereby the statistical power of neuroimaging studies. In this paper, we present a novel approach to harmonize dMRI data (the raw signal, instead of dMRI derived measures such as fractional anisotropy) using rotation invariant spherical harmonic (RISH) features embedded within a multi-modal image registration framework. All dMRI data sets from all sites are registered to a common template and voxel-wise differences in RISH features between sites at a group level are used to harmonize the signal in a subject-specific manner. We validate our method on diffusion data acquired from seven different sites (two GE, three Philips, and two Siemens scanners) on a group of age-matched healthy subjects. We demonstrate the efficacy of our method by statistically comparing diffusion measures such as fractional anisotropy, mean diffusivity and generalized fractional anisotropy across these sites before and after data harmonization. Validation was also done on a group oftest subjects, which were not used to "learn" the harmonization parameters. We also show results using TBSS before and after harmonization for independent validation of the proposed methodology. Using synthetic data, we show that any abnormality in diffusion measures due to disease is preserved during the harmonization process. Our experimental results demonstrate that, for nearly identical acquisition protocol across sites, scanner-specific differences in the signal can be removed using the proposed method in a model independent manner.

Inter-Vendor and Inter-Session Reliability of Diffusion Tensor Imaging: Implications for Multicenter Clinical Imaging Studies

Objective

To evaluate the inter-vendor and inter-session reliability of diffusion tensor imaging (DTI) and relevant parameters.

Materials and Methods

This prospective study included 10 healthy subjects (5 women and 5 men; age range, 25-33 years). Each subject was scanned twice using 3T magnetic resonance scanners from three different vendors at two different sites. A voxel-wise statistical analysis of diffusion data was performed using Tract-Based Spatial Statistics. Fractional anisotropy (FA), mean diffusivity (MD), and radial diffusivity (RD) values were calculated for each brain voxel using FMRIB's Diffusion Toolbox.

Results

A repeated measures analysis of variance revealed that there were no significant differences in FA values across the vendors or between sessions; however, there were significant differences in MD values between the vendors (p = 0.020). Although there were no significant differences in inter-session MD and inter-session/inter-vendor RD values, a significant group × factor interaction revealed differences in MD and RD values between the 1st and 2nd sessions conducted by the vendors (p = 0.004 and 0.006, respectively).

Conclusion

Although FA values exhibited good inter-vendor and inter-session reliability, MD and RD values did not show consistent results. Researchers using DTI should be aware of these limitations, especially when implementing DTI in multicenter studies.

White-Matter Pathways for Statistical Learning of Temporal Structures

Extracting the statistics of event streams in natural environments is critical for interpreting current events and predicting future ones. The brain is known to rapidly find structure and meaning in unfamiliar streams of sensory experience, often by mere exposure to the environment (i.e., without explicit feedback). Yet, we know little about the brain pathways that support this type of statistical learning. Here, we test whether changes in white-matter (WM) connectivity due to training relate to our ability to extract temporal regularities. By combining behavioral training and diffusion tensor imaging (DTI), we demonstrate that humans adapt to the environment's statistics as they change over time from simple repetition to probabilistic combinations. In particular, we show that learning relates to the decision strategy that individuals adopt when extracting temporal statistics. We next test for learning-dependent changes in WM connectivity and ask whether they relate to individual variability in decision strategy. Our DTI results provide evidence for dissociable WM pathways that relate to individual strategy: extracting the exact sequence statistics (i.e., matching) relates to connectivity changes between caudate and hippocampus, while selecting the most probable outcomes in a given context (i.e., maximizing) relates to connectivity changes between prefrontal, cingulate and basal ganglia (caudate, putamen) regions. Thus, our findings provide evidence for distinct cortico-striatal circuits that show learning-dependent changes of WM connectivity and support individual ability to learn behaviorally-relevant statistics.

Diffusion MRI of white matter microstructure development in childhood and adolescence: Methods, challenges and progress

Diffusion magnetic resonance imaging (dMRI) continues to grow in popularity as a useful neuroimaging method to study brain development, and longitudinal studies that track the same individuals over time are emerging. Over the last decade, seminal work using dMRI has provided new insights into the development of brain white matter (WM) microstructure, connections and networks throughout childhood and adolescence. This review provides an introduction to dMRI, both diffusion tensor imaging (DTI) and other dMRI models, as well as common acquisition and analysis approaches. We highlight the difficulties associated with ascribing these imaging measurements and their changes over time to specific underlying cellular and molecular events. We also discuss selected methodological challenges that are of particular relevance for studies of development, including critical choices related to image acquisition, image analysis, quality control assessment, and the within-subject and longitudinal reliability of dMRI measurements. Next, we review the exciting progress in the characterization and understanding of brain development that has resulted from dMRI studies in childhood and adolescence, including brief overviews and discussions of studies focusing on sex and individual differences. Finally, we outline future directions that will be beneficial to the field.

Mapping the order and pattern of brain structural MRI changes using change-point analysis in premanifest Huntington's disease

Huntington's disease (HD) is an autosomal dominant neurodegenerative disorder that progressively affects motor, cognitive, and emotional functions. Structural MRI studies have demonstrated brain atrophy beginning many years prior to clinical onset ("premanifest" period), but the order and pattern of brain structural changes have not been fully characterized. In this study, we investigated brain regional volumes and diffusion tensor imaging (DTI) measurements in premanifest HD, and we aim to determine (1) the extent of MRI changes in a large number of structures across the brain by atlas-based analysis, and (2) the initiation points of structural MRI changes in these brain regions. We adopted a novel multivariate linear regression model to detect the inflection points at which the MRI changes begin (namely, "change-points"), with respect to the CAG-age product (CAP, an indicator of extent of exposure to the effects of CAG repeat expansion). We used approximately 300 T1-weighted and DTI data from premanifest HD and control subjects in the PREDICT-HD study, with atlas-based whole brain segmentation and change-point analysis. The results indicated a distinct topology of structural MRI changes: the change-points of the volumetric measurements suggested a central-to-peripheral pattern of atrophy from the striatum to the deep white matter; and the change points of DTI measurements indicated the earliest changes in mean diffusivity in the deep white matter and posterior white matter. While interpretation needs to be cautious given the cross-sectional nature of the data, these findings suggest a spatial and temporal pattern of spread of structural changes within the HD brain. Hum Brain Mapp 38:5035-5050, 2017. © 2017 Wiley Periodicals, Inc.

Using DTI to assess white matter microstructure in cerebral small vessel disease (SVD) in multicentre studies

Diffusion tensor imaging (DTI) metrics such as fractional anisotropy (FA) and mean diffusivity (MD) have been proposed as clinical trial markers of cerebral small vessel disease (SVD) due to their associations with outcomes such as cognition. However, studies investigating this have been predominantly single-centre. As clinical trials are likely to be multisite, further studies are required to determine whether associations with cognition of similar strengths can be detected in a multicentre setting. One hundred and nine patients (mean age =68 years) with symptomatic lacunar infarction and confluent white matter hyperintensities (WMH) on MRI was recruited across six sites as part of the PRESERVE DTI substudy. After handling missing data, 3T-MRI scanning was available from five sites on five scanner models (Siemens and Philips), alongside neuropsychological and quality of life (QoL) assessments. FA median and MD peak height were extracted from DTI histogram analysis. Multiple linear regressions were performed, including normalized brain volume, WMH lesion load, and n° lacunes as covariates, to investigate the association of FA and MD with cognition and QoL. DTI metrics from all white matter were significantly associated with global cognition (standardized β =0.268), mental flexibility (β =0.306), verbal fluency (β =0.376), and Montreal Cognitive Assessment (MoCA) (β =0.273). The magnitudes of these associations were comparable with those previously reported from single-centre studies found in a systematic literature review. In this multicentre study, we confirmed associations between DTI parameters and cognition, which were similar in strength to those found in previous single-centre studies. The present study supports the use of DTI metrics as biomarkers of disease progression in multicentre studies.

Toward Precision and Reproducibility of Diffusion Tensor Imaging: A Multicenter Diffusion Phantom and Traveling Volunteer Study

BACKGROUND AND PURPOSE

Precision medicine is an approach to disease diagnosis, treatment, and prevention that relies on quantitative biomarkers that minimize the variability of individual patient measurements. The aim of this study was to assess the intersite variability after harmonization of a high-angular-resolution 3T diffusion tensor imaging protocol across 13 scanners at the 11 academic medical centers participating in the Transforming Research and Clinical Knowledge in Traumatic Brain Injury multisite study.

MATERIALS AND METHODS

Diffusion MR imaging was acquired from a novel isotropic diffusion phantom developed at the National Institute of Standards and Technology and from the brain of a traveling volunteer on thirteen 3T MR imaging scanners representing 3 major vendors (GE Healthcare, Philips Healthcare, and Siemens). Means of the DTI parameters and their coefficients of variation across scanners were calculated for each DTI metric and white matter tract.

RESULTS

For the National Institute of Standards and Technology diffusion phantom, the coefficients of variation of the apparent diffusion coefficient across the 13 scanners was <3.8% for a range of diffusivities from 0.4 to 1.1 × 10^-6 mm²/s. For the volunteer, the coefficients of variations across scanners of the 4 primary DTI metrics, each averaged over the entire white matter skeleton, were all <5%. In individual white matter tracts, large central pathways showed good reproducibility with the coefficients of variation consistently below 5%. However, smaller tracts showed more variability, with the coefficients of variation of some DTI metrics reaching 10%.

CONCLUSIONS

The results suggest the feasibility of standardizing DTI across 3T scanners from different MR imaging vendors in a large-scale neuroimaging research study.

White Matter Integrity in High-Altitude Pilots Exposed to Hypobaria

INTRODUCTION

Nonhypoxic hypobaric (low atmospheric pressure) occupational exposure, such as experienced by U.S. Air Force U-2 pilots and safety personnel operating inside altitude chambers, is associated with increased subcortical white matter hyperintensity (WMH) burden. The pathophysiological mechanisms underlying this discrete WMH change remain unknown. The objectives of this study were to demonstrate that occupational exposure to nonhypoxic hypobaria is associated with altered white matter integrity as quantified by fractional anisotropy (FA) measured using diffusion tensor imaging and relate these findings to WMH burden and neurocognitive ability.

METHODS

There were 102 U-2 pilots and 114 age- and gender-controlled, health-matched controls who underwent magnetic resonance imaging. All pilots performed neurocognitive assessment. Whole-brain and tract-wise average FA values were compared between pilots and controls, followed by comparison within pilots separated into high and low WMH burden groups. Neurocognitive measurements were used to help interpret group difference in FA values.

RESULTS

Pilots had significantly lower average FA values than controls (0.489/0.500, respectively). Regionally, pilots had higher FA values in the fronto-occipital tract where FA values positively correlated with visual-spatial performance scores (0.603/0.586, respectively). There was a trend for high burden pilots to have lower FA values than low burden pilots.

DISCUSSION

Nonhypoxic hypobaric exposure is associated with significantly lower average FA in young, healthy U-2 pilots. This suggests that recurrent hypobaric exposure causes diffuse axonal injury in addition to focal white matter changes.McGuire SA, Boone GRE, Sherman PM, Tate DF, Wood JD, Patel B, Eskandar G, Wijtenburg SA, Rowland LM, Clarke GD, Grogan PM, Sladky JH, Kochunov PV. White matter integrity in high-altitude pilots exposed to hypobaria. Aerosp Med Hum Perform. 2016; 87(12):983-988.

Brain atrophy in Alzheimer's Disease and aging

Thanks to its safety and accessibility, magnetic resonance imaging (MRI) is extensively used in clinical routine and research field, largely contributing to our understanding of the pathophysiology of neurodegenerative disorders such as Alzheimer's disease (AD). This review aims to provide a comprehensive overview of the main findings in AD and normal aging over the past twenty years, focusing on the patterns of gray and white matter changes assessed in vivo using MRI. Major progresses in the field concern the segmentation of the hippocampus with novel manual and automatic segmentation approaches, which might soon enable to assess also hippocampal subfields. Advancements in quantification of hippocampal volumetry might pave the way to its broader use as outcome marker in AD clinical trials. Patterns of cortical atrophy have been shown to accurately track disease progression and seem promising in distinguishing among AD subtypes. Disease progression has also been associated with changes in white matter tracts. Recent studies have investigated two areas often overlooked in AD, such as the striatum and basal forebrain, reporting significant atrophy, although the impact of these changes on cognition is still unclear. Future integration of different MRI modalities may further advance the field by providing more powerful biomarkers of disease onset and progression.

Design, rationale, and baseline characteristics of the randomized double-blind phase II clinical trial of ibudilast in progressive multiple sclerosis

BACKGROUND

Primary and secondary progressive multiple sclerosis (MS), collectively called progressive multiple sclerosis (PMS), is characterized by gradual progression of disability. The current anti-inflammatory treatments for MS have little or no efficacy in PMS in the absence of obvious active inflammation. Optimal biomarkers for phase II PMS trials is unknown. Ibudilast is an inhibitor of macrophage migration inhibitor factor and phosphodiesterases-4 and -10 and exhibits possible neuroprotective properties. The goals of SPRINT-MS study are to evaluate the safety and efficacy of ibudilast in PMS and to directly compare several imaging metrics for utility in PMS trials.

METHODS

SPRINT-MS is a randomized, placebo-controlled, phase II trial of ibudilast in patients with PMS. Eligible subjects were randomized 1:1 to receive either ibudilast (100mg/day) or placebo for 96weeks. Imaging is conducted every 24weeks for whole brain atrophy, magnetization transfer ratio, diffusion tensor imaging, cortical brain atrophy, and retinal nerve fiber layer thickness. Clinical outcomes include neurologic disability and patient reported quality of life. Safety assessments include laboratory testing, electrocardiography, and suicidality screening.

RESULTS

A total of 331 subjects were enrolled, of which 255 were randomized onto active study treatment. Randomized subjects were 53.7% female and mean age 55.7 (SD 7.3) years. The last subject is projected to complete the study in May 2017.

CONCLUSION

SPRINT-MS is designed to evaluate the safety and efficacy of ibudilast as a treatment for PMS while simultaneously validating five different imaging biomarkers as outcome metrics for use in future phase II proof-of-concept PMS trials.