Cross-scanner and cross-protocol diffusion MRI data harmonisation: A benchmark database and evaluation of algorithms

A multi-site, multi-modal travelling-heads resource for brain MRI harmonisation

Despite its great potential for studying the living brain, magnetic resonance imaging (MRI) can be often limited by nuisance non-biological factors, such as hardware/software differences between scanners, which can interfere with biological variability. This lack of standardisation or harmonisation between scanners hinders reproducibility and quantifiability of MRI. Towards addressing this challenge, we present one of the most comprehensive MRI harmonisation resources, based on a travelling heads paradigm; healthy volunteers scanned repeatedly across different scanners. The Oxford-Nottingham Harmonisation (ON-Harmony) resource offers data from 20 participants each scanned on six different 3 T MRI scanners from three major vendors (GE/Philips/Siemens) across five imaging sites. Each scanning session includes five imaging modalities (T1w/T2w/dMRI/rfMRI/SWI) with protocols aligned to the UK Biobank, while for about half of the participants five within-scanner repeats are additionally acquired. The 165 multi-modal scanning sessions allow mapping of different pools of variability (biological, between-scanner, within-scanner) for hundreds of MRI-derived measures. We describe the breadth of information contained in the publicly-available data and showcase their reuse potential for evaluating efficacy of harmonisation approaches.

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 to investigate brain structure. However, these studies face challenges due to variations in scanner hardware and acquisition protocols. While several methods for dMRI data harmonization exist, 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 three distinct datasets (OASIS-3, ADNI-2, and PREVENT-AD), comparing its performance to the widely used ComBat method and the more recent CovBat approach. 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 alternative methods. Qualitative assessments illustrate the desired distributional alignment across datasets, 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.

Assessing measurement consistency of a diffusion tensor imaging (DTI) quality control (QC) anisotropy phantom

OBJECTIVES

We evaluated a quality control (QC) phantom designed to mimic diffusion characteristics and white matter fiber tracts in the brain. We hypothesized that acquisition of diffusion tensor imaging (DTI) data on different vendors and over multiple repeated measures would not contribute to significant variability in calculated diffusion tensor scalar metrics such as fractional anisotropy (FA) and mean diffusivity (MD).

MATERIALS AND METHODS

The DTI QC phantom was scanned using a 32-direction DTI sequence on General Electric (GE), Siemens, and Philips 3 Tesla scanners. Motion probing gradients (MPGs) were investigated as a source of variance in our statistical design, and data were acquired on GE and Siemens scanners using GE, Siemens, and Philips vendor MPGs for 32 directions. In total, 8 repeated scans were made for each GE/Siemens combination of vendor and MPGs with 8 repeated scans on a Philips machine using its stock DTI sequence. Data were analyzed using 2-way ANOVAs to investigate repeat scan and vendor variances and 3-way ANOVAs with repeat, MPG, and vendor as factors.

RESULTS

No statistical differences (i.e., P > 0.05) were found in any DTI scalar metrics (FA, MD) or for any factor, suggesting system constancy across imaging platforms and the specified phantom's reliability and reproducibility across vendors and conditions.

DISCUSSION

A DTI QC phantom demonstrates that DTI measurements maintain their consistency across different MRI systems and can contribute to a standard that is more reliable for quantitative MRI analyses.

DDEvENet: Evidence-based ensemble learning for uncertainty-aware brain parcellation using diffusion MRI

In this study, we developed an Evidential Ensemble Neural Network based on Deep learning and Diffusion MRI, namely DDEvENet, for anatomical brain parcellation. The key innovation of DDEvENet is the design of an evidential deep learning framework to quantify predictive uncertainty at each voxel during a single inference. To do so, we design an evidence-based ensemble learning framework for uncertainty-aware parcellation to leverage the multiple dMRI parameters derived from diffusion MRI. Using DDEvENet, we obtained accurate parcellation and uncertainty estimates across different datasets from healthy and clinical populations and with different imaging acquisitions. The overall network includes five parallel subnetworks, where each is dedicated to learning the FreeSurfer parcellation for a certain diffusion MRI parameter. An evidence-based ensemble methodology is then proposed to fuse the individual outputs. We perform experimental evaluations on large-scale datasets from multiple imaging sources, including high-quality diffusion MRI data from healthy adults and clinically diffusion MRI data from participants with various brain diseases (schizophrenia, bipolar disorder, attention-deficit/hyperactivity disorder, Parkinson's disease, cerebral small vessel disease, and neurosurgical patients with brain tumors). Compared to several state-of-the-art methods, our experimental results demonstrate highly improved parcellation accuracy across the multiple testing datasets despite the differences in dMRI acquisition protocols and health conditions. Furthermore, thanks to the uncertainty estimation, our DDEvENet approach demonstrates a good ability to detect abnormal brain regions in patients with lesions that are consistent with expert-drawn results, enhancing the interpretability and reliability of the segmentation results.

Implicit neural representation of multi-shell constrained spherical deconvolution for continuous modeling of diffusion MRI

Diffusion magnetic resonance imaging (dMRI) provides insight into the micro and macro-structure of the brain. Multi-shell multi-tissue constrained spherical deconvolution (MSMT-CSD) models the underlying local fiber orientation distributions (FODs) using the dMRI signal. While generally producing high-quality FODs, MSMT-CSD is a voxel-wise method that can be impacted by noise and produce erroneous FODs. Local models also do not use the spatial correlation between neighboring voxels to increase parameter estimating power. Additionally, voxel-wise methods require interpolation at arbitrary locations outside of voxel centers. These interpolations can be computationally costly or inaccurate, depending on the method of choice. Expanding upon previous work, we apply the implicit neural representation (INR) methodology to the MSMT-CSD model. This results in an unsupervised machine-learning framework that generates a continuous representation of a given dMRI dataset. The input of the INR consists of coordinates in the volume, which produce the spherical harmonics coefficients parameterizing an FOD at any desired location. A key characteristic of our model is its ability to leverage spatial correlations in the volume, which acts as a form of regularization. We evaluate the output FODs quantitatively and qualitatively in synthetic and real dMRI datasets and compare them to existing methods.

Learning disentangled representations to harmonize connectome network measures

Purpose

Connectome network metrics are commonly regarded as fundamental properties of the brain, and their alterations have been implicated in the development of Alzheimer's disease, multiple sclerosis, and traumatic brain injury. However, these metrics are actually estimated properties through a multistage propagation from local voxel diffusion estimations, regional tractography, and region of interest mapping. These estimation processes are significantly influenced by choices specific to imaging protocols and software, producing site-wise effects.

Approach

Recent advances in disentanglement techniques offer opportunities to learn representational spaces that separate factors that cause domain shifts from intrinsic biological factors. Although these techniques have been applied in unsupervised brain anomaly detection and image-level features, their application to the unique manifold structures of connectome adjacency matrices remains unexplored. Here, we explore the conditional variational autoencoder structure for generating site-invariant representations of the connectome, allowing the harmonization of brain network measures.

Results

Focusing on the context of aging, we conducted a study involving 823 patients across two sites. This approach effectively segregates site-specific influences from biological features, aligns network measures across different domains (Cohen's D < 0.2 and Mann-Whitney U - test < 0.05 ), and maintains associations with age ( 2.71 × 10 - 02 ± 2.86 × 10 - 03 error in years) and sex ( 0.92 ± 0.02 accuracy).

Conclusions

Our findings demonstrate that using latent representations significantly harmonizes network measures and provides robust metrics for multi-site brain network analysis.

Diffusion MRI with Machine Learning

Diffusion-weighted magnetic resonance imaging (dMRI) of the brain offers unique capabilities including noninvasive probing of tissue microstructure and structural connectivity. It is widely used for clinical assessment of disease and injury, and for neuroscience research. Analyzing the dMRI data to extract useful information for medical and scientific purposes can be challenging. The dMRI measurements may suffer from strong noise and artifacts, and may exhibit high inter-session and inter-scanner variability in the data, as well as inter-subject heterogeneity in brain structure. Moreover, the relationship between measurements and the phenomena of interest can be highly complex. Recent years have witnessed increasing use of machine learning methods for dMRI analysis. This manuscript aims to assess these efforts, with a focus on methods that have addressed data preprocessing and harmonization, microstructure mapping, tractography, and white matter tract analysis. We study the main findings, strengths, and weaknesses of the existing methods and suggest topics for future research. We find that machine learning may be exceptionally suited to tackle some of the difficult tasks in dMRI analysis. However, for this to happen, several shortcomings of existing methods and critical unresolved issues need to be addressed. There is a pressing need to improve evaluation practices, to increase the availability of rich training datasets and validation benchmarks, as well as model generalizability, reliability, and explainability concerns.

Advanced Diffusion-Weighted MRI for Cancer Microstructure Assessment in Body Imaging, and Its Relationship With Histology

Diffusion-weighted magnetic resonance imaging (DW-MRI) aims to disentangle multiple biological signal sources in each imaging voxel, enabling the computation of innovative maps of tissue microstructure. DW-MRI model development has been dominated by brain applications. More recently, advanced methods with high fidelity to histology are gaining momentum in other contexts, for example, in oncological applications of body imaging, where new biomarkers are urgently needed. The objective of this article is to review the state-of-the-art of DW-MRI in body imaging (ie, not including the nervous system) in oncology, and to analyze its value as compared to reference colocalized histology measurements, given that demonstrating the histological validity of any new DW-MRI method is essential. In this article, we review the current landscape of DW-MRI techniques that extend standard apparent diffusion coefficient (ADC), describing their acquisition protocols, signal models, fitting settings, microstructural parameters, and relationship with histology. Preclinical, clinical, and in/ex vivo studies were included. The most used techniques were intravoxel incoherent motion (IVIM; 36.3% of used techniques), diffusion kurtosis imaging (DKI; 16.7%), vascular, extracellular, and restricted diffusion for cytometry in tumors (VERDICT; 13.3%), and imaging microstructural parameters using limited spectrally edited diffusion (IMPULSED; 11.7%). Another notable category of techniques relates to innovative b-tensor diffusion encoding or joint diffusion-relaxometry. The reviewed approaches provide histologically meaningful indices of cancer microstructure (eg, vascularization/cellularity) which, while not necessarily accurate numerically, may still provide useful sensitivity to microscopic pathological processes. Future work of the community should focus on improving the inter-/intra-scanner robustness, and on assessing histological validity in broader contexts. LEVEL OF EVIDENCE: NA TECHNICAL EFFICACY: Stage 2.

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.

CROSS-AGE AND CROSS-SITE DOMAIN SHIFT IMPACTS ON DEEP LEARNING-BASED WHITE MATTER FIBER ESTIMATION IN NEWBORN AND BABY BRAINS

Deep learning models have shown great promise in estimating tissue microstructure from limited diffusion magnetic resonance imaging data. However, these models face domain shift challenges when test and train data are from different scanners and protocols, or when the models are applied to data with inherent variations such as the developing brains of infants and children scanned at various ages. Several techniques have been proposed to address some of these challenges, such as data harmonization or domain adaptation in the adult brain. However, those techniques remain unexplored for the estimation of fiber orientation distribution functions in the rapidly developing brains of infants. In this work, we extensively investigate the age effect and domain shift within and across two different cohorts of 201 newborns and 165 babies using the Method of Moments and fine-tuning strategies. Our results show that reduced variations in the microstructural development of babies in comparison to newborns directly impact the deep learning models' cross-age performance. We also demonstrate that a small number of target domain samples can significantly mitigate domain shift problems.

A diffusion MRI tractography atlas for concurrent white matter mapping across Eastern and Western populations

The study of brain differences across Eastern and Western populations provides vital insights for understanding potential cultural and genetic influences on cognition and mental health. Diffusion MRI (dMRI) tractography is an important tool in assessing white matter (WM) connectivity and brain tissue microstructure across different populations. However, a comprehensive investigation into WM fiber tracts between Eastern and Western populations is challenged due to the lack of a cross-population WM atlas and the large site-specific variability of dMRI data. This study presents a dMRI tractography atlas, namely the East-West WM Atlas, for concurrent WM mapping between Eastern and Western populations and creates a large, harmonized dMRI dataset (n=306) based on the Human Connectome Project and the Chinese Human Connectome Project. The curated WM atlas, as well as subject-specific data including the harmonized dMRI data, the whole brain tractography data, and parcellated WM fiber tracts and their diffusion measures, are publicly released. This resource is a valuable addition to facilitating the exploration of brain commonalities and differences across diverse cultural backgrounds.

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.

Harmonizing T1-Weighted Images to Improve Consistency of Brain Morphology Among Different Scanner Manufacturers in Alzheimer's disease

BACKGROUND

Brain MRI scanner variability can introduce bias in measurements. Harmonizing scanner variability is crucial.

PURPOSE

To develop a harmonization method aimed at removing scanner variability, and to evaluate the consistency of results in multicenter studies.

STUDY TYPE

Retrospective.

POPULATION

Multicenter data from 170 healthy participants (males/females = 98/72; age = 73.8 ± 7.3) and 170 Alzheimer's disease patients (males/females = 98/72; age = 76.2 ± 8.5) were compared with reference data from another 340 participants.

FIELD STRENGTH/SEQUENCE

3-T, magnetization prepared rapid gradient echo and turbo field echo; 1.5-T, inversion recovery prepared fast spoiled gradient echo T1-weighted sequences.

ASSESSMENT

Gray matter (GM) brain images, obtained through segmentation of T1-weighted images, were utilized to evaluate the performance of the harmonization method using common orthogonal basis extraction (HCOBE) and four other methods (removal of artificial voxel effect by linear regression, RAVEL; Z_score; general linear model, GLM; ComBat). Linear discriminant analysis (LDA) was used to access the effectiveness of different methods in reducing scanner variability. The performance of harmonization methods in preserving GM volumes heterogeneity was evaluated by the similarity of the relationship between GM proportion and age in the reference and multicenter data. Furthermore, the consistency of the harmonized multicenter data with the reference data were evaluated based on classification results (train/test = 7/3) and brain atrophy.

STATISTICAL TESTS

Two-sample t-tests, area under the curve (AUC), and Dice coefficients were used to analyze the consistency of results from the reference and harmonized multicenter data. A P-value <0.01 was considered statistically significant.

RESULTS

HCOBE reduced the scanner variability from 0.09 before harmonization to 0.003 (ideal: 0, RAVEL/Z_score/GLM/ComBat = 0.087/0.003/0.006/0.013). GM volumes showed no significant difference (P = 0.52) between the reference and HCOBE-harmonized multicenter data. Consistency evaluation showed that AUC values of 0.95 for both reference and HCOBE-harmonized multicenter data (RAVEL/Z_score/GLM/ComBat = 0.86/0.86/0.84/0.89), and the Dice coefficient increased from 0.73 before harmonization to 0.82 (ideal: 1, RAVEL/Z_score/GLM/ComBat = 0.39/0.64/0.59/0.74).

DATA CONCLUSION

HCOBE may help to remove scanner variability and could improve the consistency of results in multicenter studies.

LEVEL OF EVIDENCE

2 TECHNICAL EFFICACY STAGE: 1.

Harmonized diffusion MRI data and white matter measures from the Adolescent Brain Cognitive Development Study

The Adolescent Brain Cognitive Development (ABCD) Study® has collected data from over 10,000 children across 21 sites, providing insights into adolescent brain development. However, site-specific scanner variability has made it challenging to use diffusion MRI (dMRI) data from this study. To address this, a dataset of harmonized and processed ABCD dMRI data (from release 3) has been created, comprising quality-controlled imaging data from 9,345 subjects, focusing exclusively on the baseline session, i.e., the first time point of the study. This resource required substantial computational time (approx. 50,000 CPU hours) for harmonization, whole-brain tractography, and white matter parcellation. The dataset includes harmonized dMRI data, 800 white matter clusters, 73 anatomically labeled white matter tracts in full and low resolution, and 804 different dMRI-derived measures per subject (72.3 TB total size). Accessible via the NIMH Data Archive, it offers a large-scale dMRI dataset for studying structural connectivity in child and adolescent neurodevelopment. Additionally, several post-harmonization experiments were conducted to demonstrate the success of the harmonization process on the ABCD dataset.

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.

Traveling Subject-Informed Harmonization Increases Reliability of Brain Diffusion Tensor and Neurite Mapping

Diffusion-weighted magnetic resonance imaging (dMRI) of brain has helped elucidate the microstructural changes of psychiatric and neurodegenerative disorders. Inconsistency between MRI models has hampered clinical application of dMRI-based metrics. Using harmonized dMRI data of 300 scans from 69 traveling subjects (TS) scanning the same individuals at multiple conditions with 13 MRI models and 2 protocols, the widely-used metrics such as diffusion tensor imaging (DTI) and neurite orientation dispersion and density imaging (NODDI) were evaluated before and after harmonization with a combined association test (ComBat) or TS-based general linear model (TS-GLM). Results showed that both ComBat and TS-GLM significantly reduced the effects of the MRI site, model, and protocol for diffusion metrics while maintaining the intersubject biological effects. The harmonization power of TS-GLM based on TS data model is more powerful than that of ComBat. In conclusion, our research demonstrated that although ComBat and TS-GLM harmonization approaches were effective at reducing the scanner effects of the site, model, and protocol for DTI and NODDI metrics in WM, they exhibited high retainability of biological effects. Therefore, we suggest that, after harmonizing DTI and NODDI metrics, a multisite study with large cohorts can accurately detect small pathological changes by retaining pathological effects.

Reproducible Abnormalities and Diagnostic Generalizability of White Matter in Alzheimer's Disease

Alzheimer's disease (AD) is associated with the impairment of white matter (WM) tracts. The current study aimed to verify the utility of WM as the neuroimaging marker of AD with multisite diffusion tensor imaging datasets [321 patients with AD, 265 patients with mild cognitive impairment (MCI), 279 normal controls (NC)], a unified pipeline, and independent site cross-validation. Automated fiber quantification was used to extract diffusion profiles along tracts. Random-effects meta-analyses showed a reproducible degeneration pattern in which fractional anisotropy significantly decreased in the AD and MCI groups compared with NC. Machine learning models using tract-based features showed good generalizability among independent site cross-validation. The diffusion metrics of the altered regions and the AD probability predicted by the models were highly correlated with cognitive ability in the AD and MCI groups. We highlighted the reproducibility and generalizability of the degeneration pattern of WM tracts in AD.

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.

versaFlow: a versatile pipeline for resolution adapted diffusion MRI processing and its application to studying the variability of the PRIME-DE database

The lack of "gold standards" in Diffusion Weighted Imaging (DWI) makes validation cumbersome. To tackle this task, studies use translational analysis where results in humans are benchmarked against findings in other species. Non-Human Primates (NHP) are particularly interesting for this, as their cytoarchitecture is closely related to humans. However, tools used for processing and analysis must be adapted and finely tuned to work well on NHP images. Here, we propose versaFlow, a modular pipeline implemented in Nextflow, designed for robustness and scalability. The pipeline is tailored to in vivo NHP DWI at any spatial resolution; it allows for maintainability and customization. Processes and workflows are implemented using cutting-edge and state-of-the-art Magnetic Resonance Imaging (MRI) processing technologies and diffusion modeling algorithms, namely Diffusion Tensor Imaging (DTI), Constrained Spherical Deconvolution (CSD), and DIstribution of Anisotropic MicrOstructural eNvironments in Diffusion-compartment imaging (DIAMOND). Using versaFlow, we provide an in-depth study of the variability of diffusion metrics computed on 32 subjects from 3 sites of the Primate Data Exchange (PRIME-DE), which contains anatomical T1-weighted (T1w) and T2-weighted (T2w) images, functional MRI (fMRI), and DWI of NHP brains. This dataset includes images acquired over a range of resolutions, using single and multi-shell gradient samplings, on multiple scanner vendors. We perform a reproducibility study of the processing of versaFlow using the Aix-Marseilles site's data, to ensure that our implementation has minimal impact on the variability observed in subsequent analyses. We report very high reproducibility for the majority of metrics; only gamma distribution parameters of DIAMOND display less reproducible behaviors, due to the absence of a mechanism to enforce a random number seed in the software we used. This should be taken into consideration when future applications are performed. We show that the PRIME-DE diffusion data exhibits a great level of variability, similar or greater than results obtained in human studies. Its usage should be done carefully to prevent instilling uncertainty in statistical analyses. This hints at a need for sufficient harmonization in acquisition protocols and for the development of robust algorithms capable of managing the variability induced in imaging due to differences in scanner models and/or vendors.

Validation of deep learning techniques for quality augmentation in diffusion MRI for clinical studies

The objective of this study is to evaluate the efficacy of deep learning (DL) techniques in improving the quality of diffusion MRI (dMRI) data in clinical applications. The study aims to determine whether the use of artificial intelligence (AI) methods in medical images may result in the loss of critical clinical information and/or the appearance of false information. To assess this, the focus was on the angular resolution of dMRI and a clinical trial was conducted on migraine, specifically between episodic and chronic migraine patients. The number of gradient directions had an impact on white matter analysis results, with statistically significant differences between groups being drastically reduced when using 21 gradient directions instead of the original 61. Fourteen teams from different institutions were tasked to use DL to enhance three diffusion metrics (FA, AD and MD) calculated from data acquired with 21 gradient directions and a b-value of 1000 s/mm2. The goal was to produce results that were comparable to those calculated from 61 gradient directions. The results were evaluated using both standard image quality metrics and Tract-Based Spatial Statistics (TBSS) to compare episodic and chronic migraine patients. The study results suggest that while most DL techniques improved the ability to detect statistical differences between groups, they also led to an increase in false positive. The results showed that there was a constant growth rate of false positives linearly proportional to the new true positives, which highlights the risk of generalization of AI-based tasks when assessing diverse clinical cohorts and training using data from a single group. The methods also showed divergent performance when replicating the original distribution of the data and some exhibited significant bias. In conclusion, extreme caution should be exercised when using AI methods for harmonization or synthesis in clinical studies when processing heterogeneous data in clinical studies, as important information may be altered, even when global metrics such as structural similarity or peak signal-to-noise ratio appear to suggest otherwise.

Image harmonization: A review of statistical and deep learning methods for removing batch effects and evaluation metrics for effective harmonization

Magnetic resonance imaging and computed tomography from multiple batches (e.g. sites, scanners, datasets, etc.) are increasingly used alongside complex downstream analyses to obtain new insights into the human brain. However, significant confounding due to batch-related technical variation, called batch effects, is present in this data; direct application of downstream analyses to the data may lead to biased results. Image harmonization methods seek to remove these batch effects and enable increased generalizability and reproducibility of downstream results. In this review, we describe and categorize current approaches in statistical and deep learning harmonization methods. We also describe current evaluation metrics used to assess harmonization methods and provide a standardized framework to evaluate newly-proposed methods for effective harmonization and preservation of biological information. Finally, we provide recommendations to end-users to advocate for more effective use of current methods and to methodologists to direct future efforts and accelerate development of the field.

Quantification of Diffusion Magnetic Resonance Imaging for Prognostic Prediction of Neonatal Hypoxic-Ischemic Encephalopathy

Neonatal hypoxic-ischemic encephalopathy (HIE) is the leading cause of acquired neonatal brain injury with the risk of developing serious neurological sequelae and death. An accurate and robust prediction of short- and long-term outcomes may provide clinicians and families with fundamental evidence for their decision-making, the design of treatment strategies, and the discussion of developmental intervention plans after discharge. Diffusion tensor imaging (DTI) is one of the most powerful neuroimaging tools with which to predict the prognosis of neonatal HIE by providing microscopic features that cannot be assessed by conventional magnetic resonance imaging (MRI). DTI provides various scalar measures that represent the properties of the tissue, such as fractional anisotropy (FA) and mean diffusivity (MD). Since the characteristics of the diffusion of water molecules represented by these measures are affected by the microscopic cellular and extracellular environment, such as the orientation of structural components and cell density, they are often used to study the normal developmental trajectory of the brain and as indicators of various tissue damage, including HIE-related pathologies, such as cytotoxic edema, vascular edema, inflammation, cell death, and Wallerian degeneration. Previous studies have demonstrated widespread alteration in DTI measurements in severe cases of HIE and more localized changes in neonates with mild-to-moderate HIE. In an attempt to establish cutoff values to predict the occurrence of neurological sequelae, MD and FA measurements in the corpus callosum, thalamus, basal ganglia, corticospinal tract, and frontal white matter have proven to have an excellent ability to predict severe neurological outcomes. In addition, a recent study has suggested that a data-driven, unbiased approach using machine learning techniques on features obtained from whole-brain image quantification may accurately predict the prognosis of HIE, including for mild-to-moderate cases. Further efforts are needed to overcome current challenges, such as MRI infrastructure, diffusion modeling methods, and data harmonization for clinical application. In addition, external validation of predictive models is essential for clinical application of DTI to prognostication.

Harmonized diffusion MRI data and white matter measures from the Adolescent Brain Cognitive Development Study

The Adolescent Brain Cognitive Development (ABCD) study has collected data from over 10,000 children across 21 sites, providing valuable insights into adolescent brain development. However, site-specific scanner variability has made it challenging to use diffusion MRI (dMRI) data from this study. To address this, a database of harmonized and processed ABCD dMRI data has been created, comprising quality-controlled imaging data from 9345 subjects. This resource required significant computational effort, taking ~50,000 CPU hours to harmonize the data, perform white matter parcellation, and run whole brain tractography. The database includes harmonized dMRI data, 800 white matter clusters, 73 anatomically labeled white matter tracts both in full-resolution (for analysis) and low-resolution (for visualization), and 804 different dMRI-derived measures per subject. It is available via the NIMH Data Archive and offers tremendous potential for scientific discoveries in structural connectivity studies of neurodevelopment in children and adolescents. Additionally, several post-harmonization experiments were conducted to demonstrate the success of the harmonization process on the ABCD dataset.

Superficial white matter across development, young adulthood, and aging: volume, thickness, and relationship with cortical features

Superficial white matter (SWM) represents a significantly understudied part of the human brain, despite comprising a large portion of brain volume and making up a majority of cortico-cortical white matter connections. Using multiple, high-quality datasets with large sample sizes (N = 2421, age range 5-100) in combination with methodological advances in tractography, we quantified features of SWM volume and thickness across the brain and across development, young adulthood, and aging. We had four primary aims: (1) characterize SWM thickness across brain regions (2) describe associations between SWM volume and age (3) describe associations between SWM thickness and age, and (4) quantify relationships between SWM thickness and cortical features. Our main findings are that (1) SWM thickness varies across the brain, with patterns robust across individuals and across the population at the region-level and vertex-level; (2) SWM volume shows unique volumetric trajectories with age that are distinct from gray matter and other white matter trajectories; (3) SWM thickness shows nonlinear cross-sectional changes across the lifespan that vary across regions; and (4) SWM thickness is associated with features of cortical thickness and curvature. For the first time, we show that SWM volume follows a similar trend as overall white matter volume, peaking at a similar time in adolescence, leveling off throughout adulthood, and decreasing with age thereafter. Notably, the relative fraction of total brain volume of SWM continuously increases with age, and consequently takes up a larger proportion of total white matter volume, unlike the other tissue types that decrease with respect to total brain volume. This study represents the first characterization of SWM features across the large portion of the lifespan and provides the background for characterizing normal aging and insight into the mechanisms associated with SWM development and decline.

Individualised profiling of white matter organisation in moderate-to-severe traumatic brain injury patients

BACKGROUND AND PURPOSE

Approximately 65% of moderate-to-severe traumatic brain injury (m-sTBI) patients present with poor long-term behavioural outcomes, which can significantly impair activities of daily living. Numerous diffusion-weighted MRI studies have linked these poor outcomes to decreased white matter integrity of several commissural tracts, association fibres and projection fibres in the brain. However, most studies have focused on group-based analyses, which are unable to deal with the substantial between-patient heterogeneity in m-sTBI. As a result, there is increasing interest and need in conducting individualised neuroimaging analyses.

MATERIALS AND METHODS

Here, we generated a detailed subject-specific characterisation of microstructural organisation of white matter tracts in 5 chronic patients with m-sTBI (29 - 49y, 2 females), presented as a proof-of-concept. We developed an imaging analysis framework using fixel-based analysis and TractLearn to determine whether the values of fibre density of white matter tracts at the individual patient level deviate from the healthy control group (n = 12, 8F, M_age = 35.7y, age range 25 - 64y).

RESULTS

Our individualised analysis revealed unique white matter profiles, confirming the heterogenous nature of m-sTBI and the need of individualised profiles to properly characterise the extent of injury. Future studies incorporating clinical data, as well as utilising larger reference samples and examining the test-retest reliability of the fixel-wise metrics are warranted.

CONCLUSIONS

Individualised profiles may assist clinicians in tracking recovery and planning personalised training programs for chronic m-sTBI patients, which is necessary to achieve optimal behavioural outcomes and improved quality of life.

Deep learning prediction of diffusion MRI data with microstructure-sensitive loss functions

Deep learning prediction of diffusion MRI (DMRI) data relies on the utilization of effective loss functions. Existing losses typically measure the signal-wise differences between the predicted and target DMRI data without considering the quality of derived diffusion scalars that are eventually utilized for quantification of tissue microstructure. Here, we propose two novel loss functions, called microstructural loss and spherical variance loss, to explicitly consider the quality of both the predicted DMRI data and derived diffusion scalars. We apply these loss functions to the prediction of multi-shell data and enhancement of angular resolution. Evaluation based on infant and adult DMRI data indicates that both microstructural loss and spherical variance loss improve the quality of derived diffusion scalars.

Machine Learning for Brain MRI Data Harmonisation: A Systematic Review

BACKGROUND

Magnetic Resonance Imaging (MRI) data collected from multiple centres can be heterogeneous due to factors such as the scanner used and the site location. To reduce this heterogeneity, the data needs to be harmonised. In recent years, machine learning (ML) has been used to solve different types of problems related to MRI data, showing great promise.

OBJECTIVE

This study explores how well various ML algorithms perform in harmonising MRI data, both implicitly and explicitly, by summarising the findings in relevant peer-reviewed articles. Furthermore, it provides guidelines for the use of current methods and identifies potential future research directions.

METHOD

This review covers articles published through PubMed, Web of Science, and IEEE databases through June 2022. Data from studies were analysed based on the criteria of Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA). Quality assessment questions were derived to assess the quality of the included publications.

RESULTS

a total of 41 articles published between 2015 and 2022 were identified and analysed. In the review, MRI data has been found to be harmonised either in an implicit (n = 21) or an explicit (n = 20) way. Three MRI modalities were identified: structural MRI (n = 28), diffusion MRI (n = 7) and functional MRI (n = 6).

CONCLUSION

Various ML techniques have been employed to harmonise different types of MRI data. There is currently a lack of consistent evaluation methods and metrics used across studies, and it is recommended that the issue be addressed in future studies. Harmonisation of MRI data using ML shows promises in improving performance for ML downstream tasks, while caution should be exercised when using ML-harmonised data for direct interpretation.

HYDI-DSI revisited: Constrained non-parametric EAP imaging without q-space re-gridding

Hybrid Diffusion Imaging (HYDI) was one of the first attempts to use multi-shell samplings of the q-space to infer diffusion properties beyond Diffusion Tensor Imaging (DTI) or High Angular Resolution Diffusion Imaging (HARDI). HYDI was intended as a flexible protocol embedding both DTI (for lower b-values) and HARDI (for higher b-values) processing, as well as Diffusion Spectrum Imaging (DSI) when the entire data set was exploited. In the latter case, the spherical sampling of the q-space is re-gridded by interpolation to a Cartesian lattice whose extent covers the range of acquired b-values, hence being acquisition-dependent. The Discrete Fourier Transform (DFT) is afterwards used to compute the corresponding Cartesian sampling of the Ensemble Average Propagator (EAP) in an entirely non-parametric way. From this lattice, diffusion markers such as the Return To Origin Probability (RTOP) or the Mean Squared Displacement (MSD) can be numerically estimated. We aim at re-formulating this scheme by means of a Fourier Transform encoding matrix that eliminates the need for q-space re-gridding at the same time it preserves the non-parametric nature of HYDI-DSI. The encoding matrix is adaptively designed at each voxel according to the underlying DTI approximation, so that an optimal sampling of the EAP can be pursued without being conditioned by the particular acquisition protocol. The estimation of the EAP is afterwards carried out as a regularized Quadratic Programming (QP) problem, which allows to impose positivity constraints that cannot be trivially embedded within the conventional HYDI-DSI. We demonstrate that the definition of the encoding matrix in the adaptive space allows to analytically (as opposed to numerically) compute several popular descriptors of diffusion with the unique source of error being the cropping of high frequency harmonics in the Fourier analysis of the attenuation signal. They include not only RTOP and MSD, but also Return to Axis/Plane Probabilities (RTAP/RTPP), which are defined in terms of specific spatial directions and are not available with the former HYDI-DSI. We report extensive experiments that suggest the benefits of our proposal in terms of accuracy, robustness and computational efficiency, especially when only standard, non-dedicated q-space samplings are available.

Advancing brain network models to reconcile functional neuroimaging and clinical research

Functional magnetic resonance imaging (fMRI) captures information on brain function beyond the anatomical alterations that are traditionally visually examined by neuroradiologists. However, the fMRI signals are complex in addition to being noisy, so fMRI still faces limitations for clinical applications. Here we review methods that have been proposed as potential solutions so far, namely statistical, biophysical and decoding models, with their strengths and weaknesses. We especially evaluate the ability of these models to directly predict clinical variables from their parameters (predictability) and to extract clinically relevant information regarding biological mechanisms and relevant features for classification and prediction (interpretability). We then provide guidelines for useful applications and pitfalls of such fMRI-based models in a clinical research context, looking beyond the current state of the art. In particular, we argue that the clinical relevance of fMRI calls for a new generation of models for fMRI data, which combine the strengths of both biophysical and decoding models. This leads to reliable and biologically meaningful model parameters, which thus fulfills the need for simultaneous interpretability and predictability. In our view, this synergy is fundamental for the discovery of new pharmacological and interventional targets, as well as the use of models as biomarkers in neurology and psychiatry.

Contrastive semi-supervised harmonization of single-shell to multi-shell diffusion MRI

Diffusion weighted MRI (DW-MRI) harmonization is necessary for multi-site or multi-acquisition studies. Current statistical methods address the need to harmonize from one site to another, but do not simultaneously consider the use of multiple datasets which are comprised of multiple sites, acquisitions protocols, and age demographics. This work explores deep learning methods which can generalize across these variations through semi-supervised and unsupervised learning while also learning to estimate multi-shell data from single-shell data using the Multi-shell Diffusion MRI Harmonization Challenge (MUSHAC) and Baltimore Longitudinal Study on Aging (BLSA) datasets. We compare disentanglement harmonization models, which seek to encode anatomy and acquisition in separate latent spaces, and a CycleGAN harmonization model, which uses generative adversarial networks (GAN) to perform style transfer between sites, to the baseline preprocessing and to SHORE interpolation. We find that the disentanglement models achieve superior performance in harmonizing all data while at the same transforming the input data to a single target space across several diffusion metrics (fractional anisotropy, mean diffusivity, mean kurtosis, primary eigenvector).

Site effects how-to and when: An overview of retrospective techniques to accommodate site effects in multi-site neuroimaging analyses

Site differences, or systematic differences in feature distributions across multiple data-acquisition sites, are a known source of heterogeneity that may adversely affect large-scale meta- and mega-analyses of independently collected neuroimaging data. They influence nearly all multi-site imaging modalities and biomarkers, and methods to compensate for them can improve reliability and generalizability in the analysis of genetics, omics, and clinical data. The origins of statistical site effects are complex and involve both technical differences (scanner vendor, head coil, acquisition parameters, imaging processing) and differences in sample characteristics (inclusion/exclusion criteria, sample size, ancestry) between sites. In an age of expanding international consortium research, there is a growing need to disentangle technical site effects from sample characteristics of interest. Numerous statistical and machine learning methods have been developed to control for, model, or attenuate site effects - yet to date, no comprehensive review has discussed the benefits and drawbacks of each for different use cases. Here, we provide an overview of the different existing statistical and machine learning methods developed to remove unwanted site effects from independently collected neuroimaging samples. We focus on linear mixed effect models, the ComBat technique and its variants, adjustments based on image quality metrics, normative modeling, and deep learning approaches such as generative adversarial networks. For each method, we outline the statistical foundation and summarize strengths and weaknesses, including their assumptions and conditions of use. We provide information on software availability and comment on the ease of use and the applicability of these methods to different types of data. We discuss validation and comparative reports, mention caveats and provide guidance on when to use each method, depending on context and specific research questions.

Surface-based tracking for short association fibre tractography

It is estimated that in the human brain, short association fibres (SAF) represent more than half of the total white matter volume and their involvement has been implicated in a range of neurological and psychiatric conditions. This population of fibres, however, remains relatively understudied in the neuroimaging literature. Some of the challenges pertinent to the mapping of SAF include their variable anatomical course and proximity to the cortical mantle, leading to partial volume effects and potentially affecting streamline trajectory estimation. This work considers the impact of seeding and filtering strategies and choice of scanner, acquisition, data resampling to propose a whole-brain, surface-based short (≤30-40 mm) SAF tractography approach. The framework is shown to produce longer streamlines with a predilection for connecting gyri as well as high cortical coverage. We further demonstrate that certain areas of subcortical white matter become disproportionally underrepresented in diffusion-weighted MRI data with lower angular and spatial resolution and weaker diffusion weighting; however, collecting data with stronger gradients than are usually available clinically has minimal impact, making our framework translatable to data collected on commonly available hardware. Finally, the tractograms are examined using voxel- and surface-based measures of consistency, demonstrating moderate reliability, low repeatability and high between-subject variability, urging caution when streamline count-based analyses of SAF are performed.

Mapping the human connectome using diffusion MRI at 300 mT/m gradient strength: Methodological advances and scientific impact

Tremendous efforts have been made in the last decade to advance cutting-edge MRI technology in pursuit of mapping structural connectivity in the living human brain with unprecedented sensitivity and speed. The first Connectom 3T MRI scanner equipped with a 300 mT/m whole-body gradient system was installed at the Massachusetts General Hospital in 2011 and was specifically constructed as part of the Human Connectome Project. Since that time, numerous technological advances have been made to enable the broader use of the Connectom high gradient system for diffusion tractography and tissue microstructure studies and leverage its unique advantages and sensitivity to resolving macroscopic and microscopic structural information in neural tissue for clinical and neuroscientific studies. The goal of this review article is to summarize the technical developments that have emerged in the last decade to support and promote large-scale and scientific studies of the human brain using the Connectom scanner. We provide a brief historical perspective on the development of Connectom gradient technology and the efforts that led to the installation of three other Connectom 3T MRI scanners worldwide - one in the United Kingdom in Cardiff, Wales, another in continental Europe in Leipzig, Germany, and the latest in Asia in Shanghai, China. We summarize the key developments in gradient hardware and image acquisition technology that have formed the backbone of Connectom-related research efforts, including the rich array of high-sensitivity receiver coils, pulse sequences, image artifact correction strategies and data preprocessing methods needed to optimize the quality of high-gradient strength diffusion MRI data for subsequent analyses. Finally, we review the scientific impact of the Connectom MRI scanner, including advances in diffusion tractography, tissue microstructural imaging, ex vivo validation, and clinical investigations that have been enabled by Connectom technology. We conclude with brief insights into the unique value of strong gradients for diffusion MRI and where the field is headed in the coming years.

Cross-site harmonization of multi-shell diffusion MRI measures based on rotational invariant spherical harmonics (RISH)

Quantification methods based on the acquisition of diffusion magnetic resonance imaging (dMRI) with multiple diffusion weightings (e.g., multi-shell) are becoming increasingly applied to study the in-vivo brain. Compared to single-shell data for diffusion tensor imaging (DTI), multi-shell data allows to apply more complex models such as diffusion kurtosis imaging (DKI), which attempts to capture both diffusion hindrance and restriction effects, or biophysical models such as NODDI, which attempt to increase specificity by separating biophysical components. Because of the strong dependence of the dMRI signal on the measurement hardware, DKI and NODDI metrics show scanner and site differences, much like other dMRI metrics. These effects limit the implementation of multi-shell approaches in multicenter studies, which are needed to collect large sample sizes for robust analyses. Recently, a post-processing technique based on rotation invariant spherical harmonics (RISH) features was introduced to mitigate cross-scanner differences in DTI metrics. Unlike statistical harmonization methods, which require repeated application to every dMRI metric of choice, RISH harmonization is applied once on the raw data, and can be followed by any analysis. RISH features harmonization has been tested on DTI features but not its generalizability to harmonize multi-shell dMRI. In this work, we investigated whether performing the RISH features harmonization of multi-shell dMRI data removes cross-site differences in DKI and NODDI metrics while retaining longitudinal effects. To this end, 46 subjects underwent a longitudinal (up to 3 time points) two-shell dMRI protocol at 3 imaging sites. DKI and NODDI metrics were derived before and after harmonization and compared both at the whole brain level and at the voxel level. Then, the harmonization effects on cross-sectional and on longitudinal group differences were evaluated. RISH features averaged for each of the 3 sites exhibited prominent between-site differences in the frontal and posterior part of the brain. Statistically significant differences in fractional anisotropy, mean diffusivity and mean kurtosis were observed both at the whole brain and voxel level between all the acquisition sites before harmonization, but not after. The RISH method also proved effective to harmonize NODDI metrics, particularly in white matter. The RISH based harmonization maintained the magnitude and variance of longitudinal changes as compared to the non-harmonized data of all considered metrics. In conclusion, the application of RISH feature based harmonization to multi-shell dMRI data can be used to remove cross-site differences in DKI metrics and NODDI analyses, while retaining inherent relations between longitudinal acquisitions.

Data harmonisation for information fusion in digital healthcare: A state-of-the-art systematic review, meta-analysis and future research directions

Removing the bias and variance of multicentre data has always been a challenge in large scale digital healthcare studies, which requires the ability to integrate clinical features extracted from data acquired by different scanners and protocols to improve stability and robustness. Previous studies have described various computational approaches to fuse single modality multicentre datasets. However, these surveys rarely focused on evaluation metrics and lacked a checklist for computational data harmonisation studies. In this systematic review, we summarise the computational data harmonisation approaches for multi-modality data in the digital healthcare field, including harmonisation strategies and evaluation metrics based on different theories. In addition, a comprehensive checklist that summarises common practices for data harmonisation studies is proposed to guide researchers to report their research findings more effectively. Last but not least, flowcharts presenting possible ways for methodology and metric selection are proposed and the limitations of different methods have been surveyed for future research.

Personalizing Repetitive Transcranial Magnetic Stimulation Parameters for Depression Treatment Using Multimodal Neuroimaging

Repetitive transcranial magnetic stimulation (rTMS) is a tool that can be used to administer treatment for neuropsychiatric disorders such as major depressive disorder, although the clinical efficacy is still rather modest. Overly general stimulation protocols that consider neither patient-specific depression symptomology nor individualized brain characteristics, such as anatomy or structural and functional connections, may be the cause of the high inter- and intraindividual variability in rTMS clinical responses. Multimodal neuroimaging can provide the necessary insights into individual brain characteristics and can therefore be used to personalize rTMS parameters. Optimal coil positioning should include a three-step process: 1) identify the optimal (indirect) target area based on the exact symptom pattern of the patient; 2) derive the cortical (direct) target location based on functional and/or structural connectomes derived from functional and diffusion magnetic resonance imaging data; and 3) determine the ideal coil position by computational modeling, such that the electric field distribution overlaps with the cortical target. These TMS-induced electric field simulations, derived from anatomical and diffusion magnetic resonance imaging data, can be further applied to compute optimal stimulation intensities. In addition to magnetic resonance imaging, electroencephalography can provide complementary information regarding the ongoing brain oscillations. This information can be used to determine the optimal timing and frequency of the stimuli. The heightened benefits of these personalized stimulation approaches are logically reasoned, but speculative. Randomized clinical trials will be required to compare clinical responses from standard rTMS protocols to personalized protocols. Ultimately, an optimized clinical response may result from precision protocols derived from combinations of personalized stimulation parameters.

Aging and white matter microstructure and macrostructure: a longitudinal multi-site diffusion MRI study of 1218 participants

Quantifying the microstructural and macrostructural geometrical features of the human brain's connections is necessary for understanding normal aging and disease. Here, we examine brain white matter diffusion magnetic resonance imaging data from one cross-sectional and two longitudinal data sets totaling in 1218 subjects and 2459 sessions of people aged 50-97 years. Data was drawn from well-established cohorts, including the Baltimore Longitudinal Study of Aging data set, Cambridge Centre for Ageing Neuroscience data set, and the Vanderbilt Memory & Aging Project. Quantifying 4 microstructural features and, for the first time, 11 macrostructure-based features of volume, area, and length across 120 white matter pathways, we apply linear mixed effect modeling to investigate changes in pathway-specific features over time, and document large age associations within white matter. Conventional diffusion tensor microstructure indices are the most age-sensitive measures, with positive age associations for diffusivities and negative age associations with anisotropies, with similar patterns observed across all pathways. Similarly, pathway shape measures also change with age, with negative age associations for most length, surface area, and volume-based features. A particularly novel finding of this study is that while trends were homogeneous throughout the brain for microstructure features, macrostructural features demonstrated heterogeneity across pathways, whereby several projection, thalamic, and commissural tracts exhibited more decline with age compared to association and limbic tracts. The findings from this large-scale study provide a comprehensive overview of the age-related decline in white matter and demonstrate that macrostructural features may be more sensitive to heterogeneous white matter decline. Therefore, leveraging macrostructural features may be useful for studying aging and could facilitate comparisons in a variety of diseases or abnormal conditions.

Cortico-amygdalar connectivity and externalizing/internalizing behavior in children with neurodevelopmental disorders

Background

Externalizing and internalizing behaviors contribute to clinical impairment in children with neurodevelopmental disorders (NDDs). Although associations between externalizing or internalizing behaviors and cortico-amygdalar connectivity have been found in clinical and non-clinical pediatric samples, no previous study has examined whether similar shared associations are present across children with different NDDs.

Methods

Multi-modal neuroimaging and behavioral data from the Province of Ontario Neurodevelopmental Disorders (POND) Network were used. POND participants aged 6–18 years with a primary diagnosis of autism spectrum disorder (ASD), attention-deficit/hyperactivity disorder (ADHD) or obsessive–compulsive disorder (OCD), as well as typically developing children (TDC) with T1-weighted, resting-state fMRI or diffusion weighted imaging (DWI) and parent-report Child Behavioral Checklist (CBCL) data available, were analyzed (total n = 346). Associations between externalizing or internalizing behavior and cortico-amygdalar structural and functional connectivity indices were examined using linear regressions, controlling for age, gender, and image-modality specific covariates. Behavior-by-diagnosis interaction effects were also examined.

Results

No significant linear associations (or diagnosis-by-behavior interaction effects) were found between CBCL-measured externalizing or internalizing behaviors and any of the connectivity indices examined. Post-hoc bootstrapping analyses indicated stability and reliability of these null results.

Conclusions

The current study provides evidence towards an absence of a shared linear relationship between internalizing or externalizing behaviors and cortico-amygdalar connectivity properties across a transdiagnostic sample of children with different primary NDD diagnoses and TDC. Different methodological approaches, including incorporation of multi-dimensional behavioral data (e.g., task-based fMRI) or clustering approaches may be needed to clarify complex brain-behavior relationships relevant to externalizing/internalizing behaviors in heterogeneous clinical NDD populations.

Supplementary Information

The online version contains supplementary material available at 10.1007/s00429-022-02483-0.

What's new and what's next in diffusion MRI preprocessing

Diffusion MRI (dMRI) provides invaluable information for the study of tissue microstructure and brain connectivity, but suffers from a range of imaging artifacts that greatly challenge the analysis of results and their interpretability if not appropriately accounted for. This review will cover dMRI artifacts and preprocessing steps, some of which have not typically been considered in existing pipelines or reviews, or have only gained attention in recent years: brain/skull extraction, B-matrix incompatibilities w.r.t the imaging data, signal drift, Gibbs ringing, noise distribution bias, denoising, between- and within-volumes motion, eddy currents, outliers, susceptibility distortions, EPI Nyquist ghosts, gradient deviations, B1 bias fields, and spatial normalization. The focus will be on "what's new" since the notable advances prior to and brought by the Human Connectome Project (HCP), as presented in the predecessing issue on "Mapping the Connectome" in 2013. In addition to the development of novel strategies for dMRI preprocessing, exciting progress has been made in the availability of open source tools and reproducible pipelines, databases and simulation tools for the evaluation of preprocessing steps, and automated quality control frameworks, amongst others. Finally, this review will consider practical considerations and our view on "what's next" in dMRI preprocessing.

Signatures of life course socioeconomic conditions in brain anatomy

Socioeconomic status (SES) plays a significant role in health and disease. At the same time, early-life conditions affect neural function and structure, suggesting the brain may be a conduit for the biological embedding of SES. Here, we investigate the brain anatomy signatures of SES in a large-scale population cohort aged 45-85 years. We assess both gray matter morphometry and tissue properties indicative of myelin content. Higher life course SES is associated with increased volume in several brain regions, including postcentral and temporal gyri, cuneus, and cerebellum. We observe more widespread volume differences and higher myelin content in the sensorimotor network but lower myelin content in the temporal lobe associated with childhood SES. Crucially, childhood SES differences persisted in adult brains even after controlling for adult SES, highlighting the unique contribution of early-life conditions to brain anatomy, independent of later changes in SES. These findings inform on the biological underpinnings of social inequality, particularly as they pertain to early-life conditions.

Clinical applications of magnetic resonance imaging based functional and structural connectivity

Advances in computational neuroimaging techniques have expanded the armamentarium of imaging tools available for clinical applications in clinical neuroscience. Non-invasive, in vivo brain MRI structural and functional network mapping has been used to identify therapeutic targets, define eloquent brain regions to preserve, and gain insight into pathological processes and treatments as well as prognostic biomarkers. These tools have the real potential to inform patient-specific treatment strategies. Nevertheless, a realistic appraisal of clinical utility is needed that balances the growing excitement and interest in the field with important limitations associated with these techniques. Quality of the raw data, minutiae of the processing methodology, and the statistical models applied can all impact on the results and their interpretation. A lack of standardization in data acquisition and processing has also resulted in issues with reproducibility. This limitation has had a direct impact on the reliability of these tools and ultimately, confidence in their clinical use. Advances in MRI technology and computational power as well as automation and standardization of processing methods, including machine learning approaches, may help address some of these issues and make these tools more reliable in clinical use. In this review, we will highlight the current clinical uses of MRI connectomics in the diagnosis and treatment of neurological disorders; balancing emerging applications and technologies with limitations of connectivity analytic approaches to present an encompassing and appropriate perspective.

Diffusion MRI harmonization enables joint-analysis of multicentre data of patients with cerebral small vessel disease

OBJECTIVES

Acquisition-related differences in diffusion magnetic resonance imaging (dMRI) hamper pooling of multicentre data to achieve large sample sizes. A promising solution is to harmonize the raw diffusion signal using rotation invariant spherical harmonic (RISH) features, but this has not been tested in elderly subjects. Here we aimed to establish if RISH harmonization effectively removes acquisition-related differences in multicentre dMRI of elderly subjects with cerebral small vessel disease (SVD), while preserving sensitivity to disease effects.

METHODS

Five cohorts of patients with SVD (N = 397) and elderly controls (N = 175) with 3 Tesla MRI on different systems were included. First, to establish effectiveness of harmonization, the RISH method was trained with data of 13 to 15 age and sex-matched controls from each site. Fractional anisotropy (FA) and mean diffusivity (MD) were compared in matched controls between sites using tract-based spatial statistics (TBSS) and voxel-wise analysis, before and after harmonization. Second, to assess sensitivity to disease effects, we examined whether the contrast (effect sizes of FA, MD and peak width of skeletonized MD - PSMD) between patients and controls within each site remained unaffected by harmonization. Finally, we evaluated the association between white matter hyperintensity (WMH) burden, FA, MD and PSMD using linear regression analyses both within individual cohorts as well as with pooled scans from multiple sites, before and after harmonization.

RESULTS

Before harmonization, significant differences in FA and MD were observed between matched controls of different sites (p < 0.05). After harmonization these site-differences were removed. Within each site, RISH harmonization did not alter the effect sizes of FA, MD and PSMD between patients and controls (relative change in Cohen's d = 4 %) nor the strength of association with WMH volume (relative change in R2 = 2.8 %). After harmonization, patient data of all sites could be aggregated in a single analysis to infer the association between WMH volume and FA (R2 = 0.62), MD (R2 = 0.64), and PSMD (R2 = 0.60).

CONCLUSIONS

We showed that RISH harmonization effectively removes acquisition-related differences in dMRI of elderly subjects while preserving sensitivity to SVD-related effects. This study provides proof of concept for future multicentre SVD studies with pooled datasets.

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.

Same Brain, Different Look?-The Impact of Scanner, Sequence and Preprocessing on Diffusion Imaging Outcome Parameters

In clinical diagnostics and longitudinal studies, the reproducibility of MRI assessments is of high importance in order to detect pathological changes, but developments in MRI hard- and software often outrun extended periods of data acquisition and analysis. This could potentially introduce artefactual changes or mask pathological alterations. However, if and how changes of MRI hardware, scanning protocols or preprocessing software affect complex neuroimaging outcomes from, e.g., diffusion weighted imaging (DWI) remains largely understudied. We therefore compared DWI outcomes and artefact severity of 121 healthy participants (age range 19–54 years) who underwent two matched DWI protocols (Siemens product and Center for Magnetic Resonance Research sequence) at two sites (Siemens 3T Magnetom Verio and Skyra^fit). After different preprocessing steps, fractional anisotropy (FA) and mean diffusivity (MD) maps, obtained by tensor fitting, were processed with tract-based spatial statistics (TBSS). Inter-scanner and inter-sequence variability of skeletonised FA values reached up to 5% and differed largely in magnitude and direction across the brain. Skeletonised MD values differed up to 14% between scanners. We here demonstrate that DTI outcome measures strongly depend on imaging site and software, and that these biases vary between brain regions. These regionally inhomogeneous biases may exceed and considerably confound physiological effects such as ageing, highlighting the need to harmonise data acquisition and analysis. Future studies thus need to implement novel strategies to augment neuroimaging data reliability and replicability.

Cross-scanner reproducibility and harmonization of a diffusion MRI structural brain network: A traveling subject study of multi-b acquisition

Characterization of brain networks by diffusion MRI (dMRI) has rapidly evolved, and there are ongoing movements toward data sharing and multi-center studies. To extract meaningful information from multi-center data, methods to correct for the bias caused by scanner differences, that is, harmonization, are urgently needed. In this work, we report the cross-scanner differences in structural network analyses using data from nine traveling subjects (four males and five females, 21-49 years-old) who underwent scanning using four 3T scanners (public database available from the Brain/MINDS Beyond Human Brain MRI project (http://mriportal.umin.jp/)). The reliability and reproducibility were compared to those of data from another set of four subjects (all males, 29-42 years-old) who underwent scan-rescan (interval, 105-147 days) with the same scanner as well as scan-rescan data from the Human Connectome Project database. The results demonstrated that the reliability of the edge weights and graph theory metrics was lower for data including different scanners, compared to the scan-rescan with the same scanner. Besides, systematic differences between scanners were observed, indicating the risk of bias in comparing networks obtained from different scanners directly. We further demonstrate that it is feasible to reduce inter-scanner variabilities while preserving the inter-subject differences among healthy individuals by modeling the scanner effects at the level of network matrices, when traveling-subject data are available for calibration between scanners. The present data and results are expected to serve as a basis for developing and evaluating novel harmonization methods.

On the down-sampling of diffusion MRI data along the angular dimension

BACKGROUND

It has been established that the diffusion gradient directions in diffusion MRI should be uniformly distributed in 3D spherical space, so that orientation-dependent diffusion properties (e.g., fractional anisotropy or FA) can be properly quantified. Sometimes the acquired data need to be down-sampled along the angular dimension before computing diffusion properties (e.g., to exclude data points corrupted by motion artifact; to harmonize data obtained with different protocols). It is important to quantitatively assess the impact of data down-sampling on measurement of diffusion properties.

MATERIALS AND METHODS

Here we report 1) a numerical procedure for down-sampling diffusion MRI (e.g., for data harmonization), and 2) a spatial uniformity index of diffusion directions, aiming to predict the quality of the chosen down-sampling schemes (e.g., from data harmonization; or rejection of motion corrupted data points). We quantitatively evaluated human diffusion MRI data, which were down-sampled from 64 or 60 diffusion gradient directions to 30 directions, in terms of their 1) FA value accuracy (using fully-sampled data as the ground truth), 2) FA fitting residuals, and 3) spatial uniformity indices.

RESULTS

Our experimental data show that the proposed spatial uniformity index is correlated with errors in FA obtained from down-sampled diffusion MRI data. The FA fitting residuals that are typically used to assess diffusion MRI quality are not correlated with either FA errors or spatial uniformity index.

CONCLUSIONS

These results suggest that the spatial uniformity index could be more valuable in assessing quality of down-sampled diffusion MRI data, as compared with FA fitting residual measures. We expect that our implemented software procedure should prove valuable for 1) guiding data harmonization for multi-site diffusion MRI studies, and 2) assessing the impact of rejecting motion corrupted data points on the accuracy of diffusion measures.

Detecting microstructural deviations in individuals with deep diffusion MRI tractometry

Most diffusion magnetic resonance imaging studies of disease rely on statistical comparisons between large groups of patients and healthy participants to infer altered tissue states in the brain; however, clinical heterogeneity can greatly challenge their discriminative power. There is currently an unmet need to move away from the current approach of group-wise comparisons to methods with the sensitivity to detect altered tissue states at the individual level. This would ultimately enable the early detection and interpretation of microstructural abnormalities in individual patients, an important step towards personalized medicine in translational imaging. To this end, Detect was developed to advance diffusion magnetic resonance imaging tractometry towards single-patient analysis. By operating on the manifold of white-matter pathways and learning normative microstructural features, our framework captures idiosyncrasies in patterns along white-matter pathways. Our approach paves the way from traditional group-based comparisons to true personalized radiology, taking microstructural imaging from the bench to the bedside.

Making Radiomics More Reproducible across Scanner and Imaging Protocol Variations: A Review of Harmonization Methods

Radiomics converts medical images into mineable data via a high-throughput extraction of quantitative features used for clinical decision support. However, these radiomic features are susceptible to variation across scanners, acquisition protocols, and reconstruction settings. Various investigations have assessed the reproducibility and validation of radiomic features across these discrepancies. In this narrative review, we combine systematic keyword searches with prior domain knowledge to discuss various harmonization solutions to make the radiomic features more reproducible across various scanners and protocol settings. Different harmonization solutions are discussed and divided into two main categories: image domain and feature domain. The image domain category comprises methods such as the standardization of image acquisition, post-processing of raw sensor-level image data, data augmentation techniques, and style transfer. The feature domain category consists of methods such as the identification of reproducible features and normalization techniques such as statistical normalization, intensity harmonization, ComBat and its derivatives, and normalization using deep learning. We also reflect upon the importance of deep learning solutions for addressing variability across multi-centric radiomic studies especially using generative adversarial networks (GANs), neural style transfer (NST) techniques, or a combination of both. We cover a broader range of methods especially GANs and NST methods in more detail than previous reviews.

White matter microstructure in youth at risk for serious mental illness: A comparative analysis

Identifying biomarkers of serious mental illness, such as altered white matter microstructure, can aid in early diagnosis and treatment. White matter microstructure was assessed using constrained spherical deconvolution of diffusion imaging data in a sample of 219 youth (age 12-25 years, 64.84% female) across 8 sites. Participants were classified as healthy controls (HC; n = 47), familial risk for serious mental illness (n = 31), mild-symptoms (n = 37), attenuated syndromes (n = 66), or discrete disorder (n = 38) based on clinical assessments. Fractional anisotropy (FA) and mean diffusivity (MD) values were derived for the whole brain white matter, forceps minor, anterior cingulate, anterior thalamic radiations (ATR), inferior fronto-occipital fasciculus, superior longitudinal fasciculus (SLF), and uncinate fasciculus (UF). Linear mixed effects models showed a significant effect of age on MD of the left ATR, left SLF, and left UF, and a significant effect of group on FA for all tracts examined. For most tracts, the discrete disorder group had significantly lower FA than other groups, and the attenuated syndromes group had higher FA compared to HC, with few differences between the remaining groups. White matter differences in MDD are most evident in individuals following illness onset, as few significant differences were observed in the risk phase.