Volumetric Analysis from a Harmonized Multisite Brain MRI Study of a Single Subject with Multiple Sclerosis

The role of AI for MRI-analysis in multiple sclerosis-A brief overview

Magnetic resonance imaging (MRI) has played a crucial role in the diagnosis, monitoring and treatment optimization of multiple sclerosis (MS). It is an essential component of current diagnostic criteria for its ability to non-invasively visualize both lesional and non-lesional pathology. Nevertheless, modern day usage of MRI in the clinic is limited by lengthy protocols, error-prone procedures for identifying disease markers (e.g., lesions), and the limited predictive value of existing imaging biomarkers for key disability outcomes. Recent advances in artificial intelligence (AI) have underscored the potential for AI to not only improve, but also transform how MRI is being used in MS. In this short review, we explore the role of AI in MS applications that span the entire life-cycle of an MRI image, from data collection, to lesion segmentation, detection, and volumetry, and finally to downstream clinical and scientific tasks. We conclude with a discussion on promising future directions.

Within-person changes in objectively measured activity levels as a predictor of brain atrophy in multiple sclerosis

Objective

The evaluate the association between changes in accelerometry-derived activity patterns and brain atrophy in people with multiple sclerosis (PwMS).

Methods

We included PwMS aged ≥40 years with approximately annual brain MRI who wore GT9X Actigraph accelerometers every three months over two years. Functional principal components analysis (fPCA) summarized overall activity and timing. Additional indices included total and 2-hour specific activity, sedentary time, and circadian rhythm parameters. Whole brain segmentation used SLANT-CRUISE. Generalized estimating equations quantifying between- and within-person effects modeled associations between accelerometry changes and MRI outcomes, adjusting for demographic and clinical factors.

Results

We included 233 PwMS (mean age: 54.4 years, SD: 8.6, 30% male) who wore accelerometers an average of 6.3 times over 58 days across two years. fPCA showed within-person increases in the first fPC, representing low nighttime and high morning activity, were associated with slower brain atrophy (per 1 SD increase: 0.24%; 95% CI: 0.10, 0.40; p=0.0009). Similarly, a 10% increase in 8:00-10:00 AM activity was associated with 0.49% higher whole brain volume (95% CI: 0.19, 0.79; p=0.001) over time, while increased nighttime activity (0:00-2:00) was linked to -0.28% brain volume loss (95% CI: -0.48, -0.08; p=0.007). Higher moderate-to-vigorous activity and daytime activity were associated with greater brain volume preservation longitudinally.

Interpretation

Changes in activity patterns, particularly increased nighttime activity and reduced morning activity, are linked to brain atrophy in PwMS. Accelerometry offers a scalable, sensitive method for tracking MS progression and may be beneficial as a recruitment or outcome measure in trials.

IGUANe: A 3D generalizable CycleGAN for multicenter harmonization of brain MR images

In MRI studies, the aggregation of imaging data from multiple acquisition sites enhances sample size but may introduce site-related variabilities that hinder consistency in subsequent analyses. Deep learning methods for image translation have emerged as a solution for harmonizing MR images across sites. In this study, we introduce IGUANe (Image Generation with Unified Adversarial Networks), an original 3D model that leverages the strengths of domain translation and straightforward application of style transfer methods for multicenter brain MR image harmonization. IGUANe extends CycleGAN by integrating an arbitrary number of domains for training through a many-to-one architecture. The framework based on domain pairs enables the implementation of sampling strategies that prevent confusion between site-related and biological variabilities. During inference, the model can be applied to any image, even from an unknown acquisition site, making it a universal generator for harmonization. Trained on a dataset comprising T1-weighted images from 11 different scanners, IGUANe was evaluated on data from unseen sites. The assessments included the transformation of MR images with traveling subjects, the preservation of pairwise distances between MR images within domains, the evolution of volumetric patterns related to age and Alzheimer's disease (AD), and the performance in age regression and patient classification tasks. Comparisons with other harmonization and normalization methods suggest that IGUANe better preserves individual information in MR images and is more suitable for maintaining and reinforcing variabilities related to age and AD. Future studies may further assess IGUANe in other multicenter contexts, either using the same model or retraining it for applications to different image modalities. Codes and the trained IGUANe model are available at https://github.com/RocaVincent/iguane_harmonization.git.

Brain atrophy assessment in multiple sclerosis: technical- and subject-related barriers for translation to real-world application in individual subjects

INTRODUCTION

Brain atrophy is a well-established MRI outcome for predicting clinical progression and monitoring treatment response in persons with multiple sclerosis (pwMS) at the group level. Despite the important progress made, the translation of brain atrophy assessment into clinical practice faces several challenges.

AREAS COVERED

In this review, the authors discuss technical- and subject-related barriers for implementing brain atrophy assessment as part of the clinical routine at the individual level. Substantial progress has been made to understand and mitigate technical barriers behind MRI acquisition. Numerous research and commercial segmentation techniques for volume estimation are available and technically validated, but their clinical value has not been fully established. A systematic assessment of subject-related barriers, which include genetic, environmental, biological, lifestyle, comorbidity, and aging confounders, is critical for the interpretation of brain atrophy measures at the individual subject level. Educating both medical providers and pwMS will help better clarify the benefits and limitations of assessing brain atrophy for disease monitoring and prognosis.

EXPERT OPINION

Integrating brain atrophy assessment into clinical practice for pwMS requires overcoming technical and subject-related challenges. Advances in MRI standardization, artificial intelligence, and clinician education will facilitate this process, improving disease management and potentially reducing long-term healthcare costs.

Modeling the impact of MRI acquisition bias on structural connectomes: Harmonizing structural connectomes

One way to increase the statistical power and generalizability of neuroimaging studies is to collect data at multiple sites or merge multiple cohorts. However, this usually comes with site-related biases due to the heterogeneity of scanners and acquisition parameters, negatively impacting sensitivity. Brain structural connectomes are not an exception: Being derived from T1-weighted and diffusion-weighted magnetic resonance images, structural connectivity is impacted by differences in imaging protocol. Beyond minimizing acquisition parameter differences, removing bias with postprocessing is essential. In this work we create, from the exhaustive Human Connectome Project Young Adult dataset, a resampled dataset of different b-values and spatial resolutions, modeling a cohort scanned across multiple sites. After demonstrating the statistical impact of acquisition parameters on connectivity, we propose a linear regression with explicit modeling of b-value and spatial resolution, and validate its performance on separate datasets. We show that b-value and spatial resolution affect connectivity in different ways and that acquisition bias can be reduced using a linear regression informed by the acquisition parameters while retaining interindividual differences and hence boosting fingerprinting performance. We also demonstrate the generative potential of our model, and its generalization capability in an independent dataset reflective of typical acquisition practices in clinical settings.

The impact of image contrast, resolution and reader expertise on black hole identification in Multiple Sclerosis

OBJECTIVES

In the neuroradiological work-up of Multiple Sclerosis (MS), the detection of "black holes" (BH) represent an information of undeniable importance. Nevertheless, different sequences can be used in clinical practice to evaluate BH in MS. Aim of this study was to investigate the possible impact of different sequences, resolutions, and levels of expertise on the intra- and inter-rater reliability identification of BH in MS.

METHODS

Brain MRI scans of 85 MS patients (M/F = 22/63; mean age = 36.0 ± 10.2 years) were evaluated in this prospective single-center study. The acquisition protocol included a 3 mm SE-T1w sequence, a 1 mm 3D-GrE-T1w sequence from which a resliced 3 mm sequence was also obtained. Images were evaluated independently by two readers of different expertise at baseline and after a wash-out period of 30 days. The intraclass correlation coefficient (ICC) was calculated as an index of intra and inter-reader reliability.

RESULTS

For both readers, the intra-reader ICC analysis showed that the 3 mm SE-T1w and 3 mm resliced GrE-T1w images achieved an excellent performance (both with an ICC ≥ 0.95), while 1 mm 3D-GrE-T1w scans achieved a moderate one (ICC < 0.90). The inter-reader analysis showed that each of the three sequences achieved a moderate performance (all ICCs < 0.90).

CONCLUSIONS

The 1 mm 3D-GrE-T1w sequence seems to be prone to a greater intra-reader variability compared to the 3 mm SE-T1w, with this effect being driven by the higher spatial resolution of the first sequence. To ensure reliability levels comparable with the standard SE-T1w in BH count, an assessment on a 3 mm resliced GrE-T1w sequence should be recommended.

Learning multi-site harmonization of magnetic resonance images without traveling human phantoms

Harmonization improves Magn. Reson. Imaging (MRI) data consistency and is central to effective integration of diverse imaging data acquired across multiple sites. Recent deep learning techniques for harmonization are predominantly supervised in nature and hence require imaging data of the same human subjects to be acquired at multiple sites. Data collection as such requires the human subjects to travel across sites and is hence challenging, costly, and impractical, more so when sufficient sample size is needed for reliable network training. Here we show how harmonization can be achieved with a deep neural network that does not rely on traveling human phantom data. Our method disentangles site-specific appearance information and site-invariant anatomical information from images acquired at multiple sites and then employs the disentangled information to generate the image of each subject for any target site. We demonstrate with more than 6,000 multi-site T1- and T2-weighted images that our method is remarkably effective in generating images with realistic site-specific appearances without altering anatomical details. Our method allows retrospective harmonization of data in a wide range of existing modern large-scale imaging studies, conducted via different scanners and protocols, without additional data collection.

Fractal-Based Morphometrics of Glioblastoma

Morphometrics have been able to distinguish important features of glioblastoma from magnetic resonance imaging (MRI). Using morphometrics computed on segmentations of various imaging abnormalities, we show that the average and range of lacunarity and fractal dimension values across MRI slices can be prognostic for survival. We look at the repeatability of these metrics to multiple segmentations and how they are impacted by image resolution. We speak to the challenges to overcome before these metrics are included in clinical care, and the insight that they may provide.

A three-dimensional deep learning model for inter-site harmonization of structural MR images of the brain: Extensive validation with a multicenter dataset

In multicenter MRI studies, pooling the imaging data can introduce site-related variabilities and can therefore bias the subsequent analyses. To harmonize the intensity distributions of brain MR images in a multicenter dataset, unsupervised deep learning methods can be employed. Here, we developed a model based on cycle-consistent adversarial networks for the harmonization of T1-weighted brain MR images. In contrast to previous works, it was designed to process three-dimensional whole-brain images in a stable manner while optimizing computation resources. Using six different MRI datasets for healthy adults (n=1525 in total) with different acquisition parameters, we tested the model in (i) three pairwise harmonizations with site effects of various sizes, (ii) an overall harmonization of the six datasets with different age distributions, and (iii) a traveling-subject dataset. Our results for intensity distributions, brain volumes, image quality metrics and radiomic features indicated that the MRI characteristics at the various sites had been effectively homogenized. Next, brain age prediction experiments and the observed correlation between the gray-matter volume and age showed that thanks to an appropriate training strategy and despite biological differences between the dataset populations, the model reinforced biological patterns. Furthermore, radiologic analyses of the harmonized images attested to the conservation of the radiologic information in the original images. The robustness of the harmonization model (as judged with various datasets and metrics) demonstrates its potential for application in retrospective multicenter studies.

Intersite brain MRI volumetric biases persist even in a harmonized multisubject study of multiple sclerosis

BACKGROUND AND PURPOSE

Multicenter study designs involving a variety of MRI scanners have become increasingly common. However, these present the issue of biases in image-based measures due to scanner or site differences. To assess these biases, we imaged 11 volunteers with multiple sclerosis (MS) with scan and rescan data at four sites.

METHODS

Images were acquired on Siemens or Philips scanners at 3 Tesla. Automated white matter lesion detection and whole-brain, gray and white matter, and thalamic volumetry were performed, as well as expert manual delineations of T1 magnetization-prepared rapid acquisition gradient echo and T2 fluid-attenuated inversion recovery lesions. Random-effect and permutation-based nonparametric modeling was performed to assess differences in estimated volumes within and across sites.

RESULTS

Random-effect modeling demonstrated model assumption violations for most comparisons of interest. Nonparametric modeling indicated that site explained >50% of the variation for most estimated volumes. This expanded to >75% when data from both Siemens and Philips scanners were included. Permutation tests revealed significant differences between average inter- and intrasite differences in most estimated brain volumes (P < .05). The automatic activation of spine coil elements during some acquisitions resulted in a shading artifact in these images. Permutation tests revealed significant differences between thalamic volume measurements from acquisitions with and without this artifact.

CONCLUSION

Differences in brain volumetry persisted across MR scanners despite protocol harmonization. These differences were not well explained by variance component modeling; however, statistical innovations for mitigating intersite differences show promise in reducing biases in multicenter studies of MS.

Confounding Effects on the Performance of Machine Learning Analysis of Static Functional Connectivity Computed from rs-fMRI Multi-site Data

Resting-state functional magnetic resonance imaging (rs-fMRI) is a non-invasive imaging technique widely used in neuroscience to understand the functional connectivity of the human brain. While rs-fMRI multi-site data can help to understand the inner working of the brain, the data acquisition and processing of this data has many challenges. One of the challenges is the variability of the data associated with different acquisitions sites, and different MRI machines vendors. Other factors such as population heterogeneity among different sites, with variables such as age and gender of the subjects, must also be considered. Given that most of the machine-learning models are developed using these rs-fMRI multi-site data sets, the intrinsic confounding effects can adversely affect the generalizability and reliability of these computational methods, as well as the imposition of upper limits on the classification scores. This work aims to identify the phenotypic and imaging variables producing the confounding effects, as well as to control these effects. Our goal is to maximize the classification scores obtained from the machine learning analysis of the Autism Brain Imaging Data Exchange (ABIDE) rs-fMRI multi-site data. To achieve this goal, we propose novel methods of stratification to produce homogeneous sub-samples of the 17 ABIDE sites, as well as the generation of new features from the static functional connectivity values, using multiple linear regression models, ComBat harmonization models, and normalization methods. The main results obtained with our statistical models and methods are an accuracy of 76.4%, sensitivity of 82.9%, and specificity of 77.0%, which are 8.8%, 20.5%, and 7.5% above the baseline classification scores obtained from the machine learning analysis of the static functional connectivity computed from the ABIDE rs-fMRI multi-site data.

BIANCA-MS: An optimized tool for automated multiple sclerosis lesion segmentation

In this work we present BIANCA-MS, a novel tool for brain white matter lesion segmentation in multiple sclerosis (MS), able to generalize across both the wide spectrum of MRI acquisition protocols and the heterogeneity of manually labeled data. BIANCA-MS is based on the original version of BIANCA and implements two innovative elements: a harmonized setting, tested under different MRI protocols, which avoids the need to further tune algorithm parameters to each dataset; and a cleaning step developed to improve consistency in automated and manual segmentations, thus reducing unwanted variability in output segmentations and validation data. BIANCA-MS was tested on three datasets, acquired with different MRI protocols. First, we compared BIANCA-MS to other widely used tools. Second, we tested how BIANCA-MS performs in separate datasets. Finally, we evaluated BIANCA-MS performance on a pooled dataset where all MRI data were merged. We calculated the overlap using the DICE spatial similarity index (SI) as well as the number of false positive/negative clusters (nFPC/nFNC) in comparison to the manual masks processed with the cleaning step. BIANCA-MS clearly outperformed other available tools in both high- and low-resolution images and provided comparable performance across different scanning protocols, sets of modalities and image resolutions. BIANCA-MS performance on the pooled dataset (SI: 0.72 ± 0.25, nFPC: 13 ± 11, nFNC: 4 ± 8) were comparable to those achieved on each individual dataset (median across datasets SI: 0.72 ± 0.28, nFPC: 14 ± 11, nFNC: 4 ± 8). Our findings suggest that BIANCA-MS is a robust and accurate approach for automated MS lesion segmentation.

MISPEL: A supervised deep learning harmonization method for multi-scanner neuroimaging data

Large-scale data obtained from aggregation of already collected multi-site neuroimaging datasets has brought benefits such as higher statistical power, reliability, and robustness to the studies. Despite these promises from growth in sample size, substantial technical variability stemming from differences in scanner specifications exists in the aggregated data and could inadvertently bias any downstream analyses on it. Such a challenge calls for data normalization and/or harmonization frameworks, in addition to comprehensive criteria to estimate the scanner-related variability and evaluate the harmonization frameworks. In this study, we propose MISPEL (Multi-scanner Image harmonization via Structure Preserving Embedding Learning), a supervised multi-scanner harmonization method that is naturally extendable to more than two scanners. We also designed a set of criteria to investigate the scanner-related technical variability and evaluate the harmonization techniques. As an essential requirement of our criteria, we introduced a multi-scanner matched dataset of 3T T1 images across four scanners, which, to the best of our knowledge is one of the few datasets of this kind. We also investigated our evaluations using two popular segmentation frameworks: FSL and segmentation in statistical parametric mapping (SPM). Lastly, we compared MISPEL to popular methods of normalization and harmonization, namely White Stripe, RAVEL, and CALAMITI. MISPEL outperformed these methods and is promising for many other neuroimaging modalities.

The robustness of persistent homology of brain networks to data acquisition-related non-neural variability in resting state fMRI

There is increasing interest in investigating brain function based on functional connectivity networks (FCN) obtained from resting-state functional magnetic resonance imaging (fMRI). FCNs, typically obtained using measures of time series association such as Pearson's correlation, are sensitive to data acquisition parameters such as sampling period. This introduces non-neural variability in data pooled from different acquisition protocols and MRI scanners, negating the advantages of larger sample sizes in pooled data. To address this, we hypothesize that the topology or shape of brain networks must be preserved irrespective of how densely it is sampled, and metrics which capture this topology may be statistically similar across sampling periods, thereby alleviating this source of non-neural variability. Accordingly, we present an end-to-end pipeline that uses persistent homology (PH), a branch of topological data analysis, to demonstrate similarity across FCNs acquired at different temporal sampling periods. PH, as a technique, extracts topological features by capturing the network organization across all continuous threshold values, as opposed to graph theoretic methods, which fix a discrete network topology by thresholding the connectivity matrix. The extracted topological features are encoded in the form of persistent diagrams that can be compared against one another using the earth-moving metric, also popularly known as the Wasserstein distance. We extract topological features from three data cohorts, each acquired at different temporal sampling periods and demonstrate that these features are statistically the same, hence, empirically showing that PH may be robust to changes in data acquisition parameters such as sampling period.

Improving the efficiency of clinical trials in multiple sclerosis

BACKGROUND

Phase 3 clinical trials for disease-modifying therapies in relapsing-remitting multiple sclerosis (RRMS) have utilized a limited number of conventional designs with a high degree of success. However, these designs limit the types of questions that can be addressed, and the time and cost required. Moreover, trials involving people with progressive multiple sclerosis (MS) have been less successful.

OBJECTIVE

The objective of this paper is to discuss complex innovative trial designs, intermediate and composite outcomes and to improve the efficiency of trial design in MS and broaden questions that can be addressed, particularly as applied to progressive MS.

METHODS

We held an international workshop with experts in clinical trial design.

RESULTS

Recommendations include increasing the use of complex innovative designs, developing biomarkers to enrich progressive MS trial populations, prioritize intermediate outcomes for further development that target therapeutic mechanisms of action other than peripherally mediated inflammation, investigate acceptability to people with MS of data linkage for studying long-term outcomes of clinical trials, use Bayesian designs to potentially reduce sample sizes required for pediatric trials, and provide sustained funding for platform trials and registries that can support pragmatic trials.

CONCLUSION

Novel trial designs and further development of intermediate outcomes may improve clinical trial efficiency in MS and address novel therapeutic questions.

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.

Characterization of the effects of outliers on ComBat harmonization for removing inter-site data heterogeneity in multisite neuroimaging studies

Data harmonization is a key step widely used in multisite neuroimaging studies to remove inter-site heterogeneity of data distribution. However, data harmonization may even introduce additional inter-site differences in neuroimaging data if outliers are present in the data of one or more sites. It remains unclear how the presence of outliers could affect the effectiveness of data harmonization and consequently the results of analyses using harmonized data. To address this question, we generated a normal simulation dataset without outliers and a series of simulation datasets with outliers of varying properties (e.g., outlier location, outlier quantity, and outlier score) based on a real large-sample neuroimaging dataset. We first verified the effectiveness of the most commonly used ComBat harmonization method in the removal of inter-site heterogeneity using the normal simulation data, and then characterized the effects of outliers on the effectiveness of ComBat harmonization and on the results of association analyses between brain imaging-derived phenotypes and a simulated behavioral variable using the simulation datasets with outliers. We found that, although ComBat harmonization effectively removed the inter-site heterogeneity in multisite data and consequently improved the detection of the true brain-behavior relationships, the presence of outliers could damage severely the effectiveness of ComBat harmonization in the removal of data heterogeneity or even introduce extra heterogeneity in the data. Moreover, we found that the effects of outliers on the improvement of the detection of brain-behavior associations by ComBat harmonization were dependent on how such associations were assessed (i.e., by Pearson correlation or Spearman correlation), and on the outlier location, quantity, and outlier score. These findings help us better understand the influences of outliers on data harmonization and highlight the importance of detecting and removing outliers prior to data harmonization in multisite neuroimaging studies.

A comprehensive evaluation of multicentric reliability of single-subject cortical morphological networks on traveling subjects

Despite the prevalence of research on single-subject cerebral morphological networks in recent years, whether they can offer a reliable way for multicentric studies remains largely unknown. Using two multicentric datasets of traveling subjects, this work systematically examined the inter-site test-retest (TRT) reliabilities of single-subject cerebral morphological networks, and further evaluated the effects of several key factors. We found that most graph-based network measures exhibited fair to excellent reliabilities regardless of different analytical pipelines. Nevertheless, the reliabilities were affected by choices of morphological index (fractal dimension > sulcal depth > gyrification index > cortical thickness), brain parcellation (high-resolution > low-resolution), thresholding method (proportional > absolute), and network type (binarized > weighted). For the factor of similarity measure, its effects depended on the thresholding method used (absolute: Kullback-Leibler divergence > Jensen-Shannon divergence; proportional: Jensen-Shannon divergence > Kullback-Leibler divergence). Furthermore, longer data acquisition intervals and different scanner software versions significantly reduced the reliabilities. Finally, we showed that inter-site reliabilities were significantly lower than intra-site reliabilities for single-subject cerebral morphological networks. Altogether, our findings propose single-subject cerebral morphological networks as a promising approach for multicentric human connectome studies, and offer recommendations on how to determine analytical pipelines and scanning protocols for obtaining reliable results.

Multiscale functional connectivity patterns of the aging brain learned from harmonized rsfMRI data of the multi-cohort iSTAGING study

To learn multiscale functional connectivity patterns of the aging brain, we built a brain age prediction model of functional connectivity measures at seven scales on a large fMRI dataset, consisting of resting-state fMRI scans of 4186 individuals with a wide age range (22 to 97 years, with an average of 63) from five cohorts. We computed multiscale functional connectivity measures of individual subjects using a personalized functional network computational method, harmonized the functional connectivity measures of subjects from multiple datasets in order to build a functional brain age model, and finally evaluated how functional brain age gap correlated with cognitive measures of individual subjects. Our study has revealed that functional connectivity measures at multiple scales were more informative than those at any single scale for the brain age prediction, the data harmonization significantly improved the brain age prediction performance, and the data harmonization in the functional connectivity measures' tangent space worked better than in their original space. Moreover, brain age gap scores of individual subjects derived from the brain age prediction model were significantly correlated with clinical and cognitive measures. Overall, these results demonstrated that multiscale functional connectivity patterns learned from a large-scale multi-site rsfMRI dataset were informative for characterizing the aging brain and the derived brain age gap was associated with cognitive and clinical measures.

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.

Harmonization Strategies in Multicenter MRI-Based Radiomics

Radiomics analysis is a powerful tool aiming to provide diagnostic and prognostic patient information directly from images that are decoded into handcrafted features, comprising descriptors of shape, size and textural patterns. Although radiomics is gaining momentum since it holds great promise for accelerating digital diagnostics, it is susceptible to bias and variation due to numerous inter-patient factors (e.g., patient age and gender) as well as inter-scanner ones (different protocol acquisition depending on the scanner center). A variety of image and feature based harmonization methods has been developed to compensate for these effects; however, to the best of our knowledge, none of these techniques has been established as the most effective in the analysis pipeline so far. To this end, this review provides an overview of the challenges in optimizing radiomics analysis, and a concise summary of the most relevant harmonization techniques, aiming to provide a thorough guide to the radiomics harmonization process.

Prediction of the information processing speed performance in multiple sclerosis using a machine learning approach in a large multicenter magnetic resonance imaging data set

Many patients with multiple sclerosis (MS) experience information processing speed (IPS) deficits, and the Symbol Digit Modalities Test (SDMT) has been recommended as a valid screening test. Magnetic resonance imaging (MRI) has markedly improved the understanding of the mechanisms associated with cognitive deficits in MS. However, which structural MRI markers are the most closely related to cognitive performance is still unclear. We used the multicenter 3T-MRI data set of the Italian Neuroimaging Network Initiative to extract multimodal data (i.e., demographic, clinical, neuropsychological, and structural MRIs) of 540 MS patients. We aimed to assess, through machine learning techniques, the contribution of brain MRI structural volumes in the prediction of IPS deficits when combined with demographic and clinical features. We trained and tested the eXtreme Gradient Boosting (XGBoost) model following a rigorous validation scheme to obtain reliable generalization performance. We carried out a classification and a regression task based on SDMT scores feeding each model with different combinations of features. For the classification task, the model trained with thalamus, cortical gray matter, hippocampus, and lesions volumes achieved an area under the receiver operating characteristic curve of 0.74. For the regression task, the model trained with cortical gray matter and thalamus volumes, EDSS, nucleus accumbens, lesions, and putamen volumes, and age reached a mean absolute error of 0.95. In conclusion, our results confirmed that damage to cortical gray matter and relevant deep and archaic gray matter structures, such as the thalamus and hippocampus, is among the most relevant predictors of cognitive performance in MS.

A guide for researchers seeking training in retrospective data harmonization for population neuroscience studies of Alzheimer's disease and related dementias

Due to needs surrounding rigor and reproducibility, subgroup specific disease knowledge, and questions of external validity, data harmonization is an essential tool in population neuroscience of Alzheimer's disease and related dementias (ADRD). Systematic harmonization of data elements is necessary to pool information from heterogeneous samples, and such pooling allows more expansive evaluations of health disparities, more precise effect estimates, and more opportunities to discover effective prevention or treatment strategies. The key goal of this Tutorial in Population Neuroimaging Curriculum, Instruction, and Pedagogy article is to guide researchers in creating a customized population neuroscience of ADRD harmonization training plan to fit their needs or those of their mentees. We provide brief guidance for retrospective data harmonization of multiple data types in this area, including: (1) clinical and demographic, (2) neuropsychological, and (3) neuroimaging data. Core competencies and skills are reviewed, and resources are provided to fill gaps in training as well as data needs. We close with an example study in which harmonization is a critical tool. While several aspects of this tutorial focus specifically on ADRD, the concepts and resources are likely to benefit population neuroscientists working in a range of research areas.

Multisite MRI reproducibility of lateral ventricular volume using the NAIMS cooperative pilot dataset

BACKGROUND AND PURPOSE

The North American Imaging in Multiple Sclerosis (NAIMS) multisite project identified interscanner reproducibility issues with T1-based whole brain volume (WBV). Lateral ventricular volume (LVV) acquired on T2-fluid-attenuated inverse recovery (FLAIR) scans has been proposed as a robust proxy measure. Therefore, we sought to determine the relative magnitude of scanner-induced T2-FLAIR-based LVV and T1-based WBV measurement errors in relation to clinically meaningful changes.

METHODS

This was a post hoc analysis of the NAIMS pilot dataset in which a relapsing-remitting MS patient with no intrastudy clinical or radiological activity was imaged twice on seven different Siemens scanners across the United States. LVV was determined using the automated NeuroSTREAM technique on T2-FLAIR and WBV was determined with SIENAX on high-resolution T1-MPRAGE. Average LVV and WBV were measured, and absolute intrascanner and interscanner coefficients of variation (CoVs) were calculated. The variabilities were compared to previously established annual pathological and clinically meaningful cutoffs of 0.40% for WBV and of 3.51% for LVV.

RESULTS

Mean LVV across all seven scan/rescan pairs was 45.87 ± 1.15 ml. Average LVV intrascanner CoV was 1.42% and interscanner CoV was 1.78%, both smaller than the reported annualized clinically meaningful cutoff of 3.51%. In contrast, intra- and interscanner CoVs for WBV (0.99% and 1.15%) were both higher than the established cutoff of 0.40%. Individually, 1/7 intrasite and 2/7 intersite pair-wise LVV comparisons were above the 3.51% cutoff, whereas 4/7 intrasite and 7/7 intersite WBV comparisons were above the 0.40% cutoff.

CONCLUSION

Fully automated LVV segmentation has higher absolute variability than WBV, but much lower relative variability compared to clinically relevant changes, and may therefore be a meaningful proxy outcome measure of neurodegeneration.

Brain atrophy measurement over a MRI scanner change in multiple sclerosis

BACKGROUND

A change in MRI hardware impacts brain volume measurements. The aim of this study was to use MRI data from multiple sclerosis (MS) patients and healthy control subjects (HCs) to statistically model how to adjust brain atrophy measures in MS patients after a major scanner upgrade.

METHODS

We scanned 20 MS patients and 26 HCs before and three months after a major scanner upgrade (1.5 T Siemens Healthineers Magnetom Avanto to 3 T Siemens Healthineers Skyra Fit). The patient group also underwent standardized serial MRIs before and after the scanner change. Percentage whole brain volume changes (PBVC) measured by Structural Image Evaluation using Normalization of Atrophy (SIENA) in the HCs was used to estimate a corrective term based on a linear model. The factor was internally validated in HCs, and then applied to the MS group.

RESULTS

Mean PBVC during the scanner change was higher in MS than HCs (-4.1 ± 0.8 % versus -3.4 ± 0.6 %). A fixed corrective term of 3.4 (95% confidence interval: 3.13-3.67)% was estimated based on the observed average changes in HCs. Age and gender did not have a significant influence on this corrective term. After adjustment, a linear mixed effects model showed that the brain atrophy measures in MS during the scanner upgrade were not anymore associated with the scanner type (old vs new scanner; p = 0.29).

CONCLUSION

A scanner change affects brain atrophy measures in longitudinal cohorts. The inclusion of a corrective term based on changes observed in HCs helps to adjust for the known and unknown factors associated with a scanner upgrade on a group level.

AutoComBat: a generic method for harmonizing MRI-based radiomic features

The use of multicentric data is becoming essential for developing generalizable radiomic signatures. In particular, Magnetic Resonance Imaging (MRI) data used in brain oncology are often heterogeneous in terms of scanners and acquisitions, which significantly impact quantitative radiomic features. Various methods have been proposed to decrease dependency, including methods acting directly on MR images, i.e., based on the application of several preprocessing steps before feature extraction or the ComBat method, which harmonizes radiomic features themselves. The ComBat method used for radiomics may be misleading and presents some limitations, such as the need to know the labels associated with the "batch effect". In addition, a statistically representative sample is required and the applicability of a signature whose batch label is not present in the train set is not possible. This work aimed to compare a priori and a posteriori radiomic harmonization methods and propose a code adaptation to be machine learning compatible. Furthermore, we have developed AutoComBat, which aims to automatically determine the batch labels, using either MRI metadata or quality metrics as inputs of the proposed constrained clustering. A heterogeneous dataset consisting of high and low-grade gliomas coming from eight different centers was considered. The different methods were compared based on their ability to decrease relative standard deviation of radiomic features extracted from white matter and on their performance on a classification task using different machine learning models. ComBat and AutoComBat using image-derived quality metrics as inputs for batch assignment and preprocessing methods presented promising results on white matter harmonization, but with no clear consensus for all MR images. Preprocessing showed the best results on the T1w-gd images for the grading task. For T2w-flair, AutoComBat, using either metadata plus quality metrics or metadata alone as inputs, performs better than the conventional ComBat, highlighting its potential for data harmonization. Our results are MRI weighting, feature class and task dependent and require further investigations on other datasets.

MAGNIMS recommendations for harmonization of MRI data in MS multicenter studies

•

Sharing data from cooperative studies is essential to develop new biomarkers in MS.

•

Differences in MRI acquisition, analysis, storage represent a substantial constraint.

•

We review the state of the art and developments in the harmonization of MRI.

•

We provide recommendations to harmonize large MRI datasets in the MS field.

There is an increasing need of sharing harmonized data from large, cooperative studies as this is essential to develop new diagnostic and prognostic biomarkers. In the field of multiple sclerosis (MS), the issue has become of paramount importance due to the need to translate into the clinical setting some of the most recent MRI achievements. However, differences in MRI acquisition parameters, image analysis and data storage across sites, with their potential bias, represent a substantial constraint. This review focuses on the state of the art, recent technical advances, and desirable future developments of the harmonization of acquisition, analysis and storage of large-scale multicentre MRI data of MS cohorts. Huge efforts are currently being made to achieve all the requirements needed to provide harmonized MRI datasets in the MS field, as proper management of large imaging datasets is one of our greatest opportunities and challenges in the coming years. Recommendations based on these achievements will be provided here. Despite the advances that have been made, the complexity of these tasks requires further research by specialized academical centres, with dedicated technical and human resources. Such collective efforts involving different professional figures are of crucial importance to offer to MS patients a personalised management while minimizing consumption of resources.

Quantitative Structural Brain Magnetic Resonance Imaging Analyses: Methodological Overview and Application to Rett Syndrome

Congenital genetic disorders often present with neurological manifestations such as neurodevelopmental disorders, motor developmental retardation, epilepsy, and involuntary movement. Through qualitative morphometric evaluation of neuroimaging studies, remarkable structural abnormalities, such as lissencephaly, polymicrogyria, white matter lesions, and cortical tubers, have been identified in these disorders, while no structural abnormalities were identified in clinical settings in a large population. Recent advances in data analysis programs have led to significant progress in the quantitative analysis of anatomical structural magnetic resonance imaging (MRI) and diffusion-weighted MRI tractography, and these approaches have been used to investigate psychological and congenital genetic disorders. Evaluation of morphometric brain characteristics may contribute to the identification of neuroimaging biomarkers for early diagnosis and response evaluation in patients with congenital genetic diseases. This mini-review focuses on the methodologies and attempts employed to study Rett syndrome using quantitative structural brain MRI analyses, including voxel- and surface-based morphometry and diffusion-weighted MRI tractography. The mini-review aims to deepen our understanding of how neuroimaging studies are used to examine congenital genetic disorders.

Privacy-preserving harmonization via distributed ComBat

Challenges in clinical data sharing and the need to protect data privacy have led to the development and popularization of methods that do not require directly transferring patient data. In neuroimaging, integration of data across multiple institutions also introduces unwanted biases driven by scanner differences. These scanner effects have been shown by several research groups to severely affect downstream analyses. To facilitate the need of removing scanner effects in a distributed data setting, we introduce distributed ComBat, an adaptation of a popular harmonization method for multivariate data that borrows information across features. We present our fast and simple distributed algorithm and show that it yields equivalent results using data from the Alzheimer's Disease Neuroimaging Initiative. Our method enables harmonization while ensuring maximal privacy protection, thus facilitating a broad range of downstream analyses in functional and structural imaging studies.

Automated Quantification of Brain Lesion Volume From Post-trauma MR Diffusion-Weighted Images

Objectives

Determining the volume of brain lesions after trauma is challenging. Manual delineation is observer-dependent and time-consuming and cannot therefore be used in routine practice. The study aimed to evaluate the feasibility of an automated atlas-based quantification procedure (AQP) based on the detection of abnormal mean diffusivity (MD) values computed from diffusion-weighted MR images.

Methods

The performance of AQP was measured against manual delineation consensus by independent raters in two series of experiments based on: (i) realistic trauma phantoms (n = 5) where low and high MD values were assigned to healthy brain images according to the intensity, form and location of lesion observed in real TBI cases; (ii) severe TBI patients (n = 12 patients) who underwent MR imaging within 10 days after injury.

Results

In realistic TBI phantoms, no statistical differences in Dice similarity coefficient, precision and brain lesion volumes were found between AQP, the rater consensus and the ground truth lesion delineations. Similar findings were obtained when comparing AQP and manual annotations for TBI patients. The intra-class correlation coefficient between AQP and manual delineation was 0.70 in realistic phantoms and 0.92 in TBI patients. The volume of brain lesions detected in TBI patients was 59 ml (19–84 ml) (median; 25–75th centiles).

Conclusions

Our results support the feasibility of using an automated quantification procedure to determine, with similar accuracy to manual delineation, the volume of low and high MD brain lesions after trauma, and thus allow the determination of the type and volume of edematous brain lesions. This approach had comparable performance with manual delineation by a panel of experts. It will be tested in a large cohort of patients enrolled in the multicenter OxyTC trial (NCT02754063).

Deep Learning in Large and Multi-Site Structural Brain MR Imaging Datasets

Large, multi-site, heterogeneous brain imaging datasets are increasingly required for the training, validation, and testing of advanced deep learning (DL)-based automated tools, including structural magnetic resonance (MR) image-based diagnostic and treatment monitoring approaches. When assembling a number of smaller datasets to form a larger dataset, understanding the underlying variability between different acquisition and processing protocols across the aggregated dataset (termed “batch effects”) is critical. The presence of variation in the training dataset is important as it more closely reflects the true underlying data distribution and, thus, may enhance the overall generalizability of the tool. However, the impact of batch effects must be carefully evaluated in order to avoid undesirable effects that, for example, may reduce performance measures. Batch effects can result from many sources, including differences in acquisition equipment, imaging technique and parameters, as well as applied processing methodologies. Their impact, both beneficial and adversarial, must be considered when developing tools to ensure that their outputs are related to the proposed clinical or research question (i.e., actual disease-related or pathological changes) and are not simply due to the peculiarities of underlying batch effects in the aggregated dataset. We reviewed applications of DL in structural brain MR imaging that aggregated images from neuroimaging datasets, typically acquired at multiple sites. We examined datasets containing both healthy control participants and patients that were acquired using varying acquisition protocols. First, we discussed issues around Data Access and enumerated the key characteristics of some commonly used publicly available brain datasets. Then we reviewed methods for correcting batch effects by exploring the two main classes of approaches: Data Harmonization that uses data standardization, quality control protocols or other similar algorithms and procedures to explicitly understand and minimize unwanted batch effects; and Domain Adaptation that develops DL tools that implicitly handle the batch effects by using approaches to achieve reliable and robust results. In this narrative review, we highlighted the advantages and disadvantages of both classes of DL approaches, and described key challenges to be addressed in future studies.

A multi-scanner neuroimaging data harmonization using RAVEL and ComBat

Modern neuroimaging studies frequently combine data collected from multiple scanners and experimental conditions. Such data often contain substantial technical variability associated with image intensity scale (image intensity scales are not the same in different images) and scanner effects (images obtained from different scanners contain substantial technical biases). Here we evaluate and compare results of data analysis methods without any data transformation (RAW), with intensity normalization using RAVEL, with regional harmonization methods using ComBat, and a combination of RAVEL and ComBat. Methods are evaluated on a unique sample of 16 study participants who were scanned on both 1.5T and 3T scanners a few months apart. Neuroradiological evaluation was conducted for 7 different regions of interest (ROI's) pertinent to Alzheimer's disease (AD). Cortical measures and results indicate that: (1) RAVEL substantially improved the reproducibility of image intensities; (2) ComBat is preferred over RAVEL and the RAVEL-ComBat combination in terms of regional level harmonization due to more consistent harmonization across subjects and image-derived measures; (3) RAVEL and ComBat substantially reduced bias compared to analysis of RAW images, but RAVEL also resulted in larger variance; and (4) the larger root mean square deviation (RMSD) of RAVEL compared to ComBat is due mainly to its larger variance.

Inter- and Intra-Scanner Variability of Automated Brain Volumetry on Three Magnetic Resonance Imaging Systems in Alzheimer's Disease and Controls

Magnetic Resonance Imaging (MRI) has become part of the clinical routine for diagnosing neurodegenerative disorders. Since acquisitions are performed at multiple centers using multiple imaging systems, detailed analysis of brain volumetry differences between MRI systems and scan-rescan acquisitions can provide valuable information to correct for different MRI scanner effects in multi-center longitudinal studies. To this end, five healthy controls and five patients belonging to various stages of the AD continuum underwent brain MRI acquisitions on three different MRI systems (Philips Achieva dStream 1.5T, Philips Ingenia 3T, and GE Discovery MR750w 3T) with harmonized scan parameters. Each participant underwent two subsequent MRI scans per imaging system, repeated on three different MRI systems within 2 h. Brain volumes computed by icobrain dm (v5.0) were analyzed using absolute and percentual volume differences, Dice similarity (DSC) and intraclass correlation coefficients, and coefficients of variation (CV). Harmonized scans obtained with different scanners of the same manufacturer had a measurement error closer to the intra-scanner performance. The gap between intra- and inter-scanner comparisons grew when comparing scans from different manufacturers. This was observed at image level (image contrast, similarity, and geometry) and translated into a higher variability of automated brain volumetry. Mixed effects modeling revealed a significant effect of scanner type on some brain volumes, and of the scanner combination on DSC. The study concluded a good intra- and inter-scanner reproducibility, as illustrated by an average intra-scanner (inter-scanner) CV below 2% (5%) and an excellent overlap of brain structure segmentation (mean DSC > 0.88).

Multi-scanner Harmonization of Paired Neuroimaging Data via Structure Preserving Embedding Learning

Combining datasets from multiple sites/scanners has been becoming increasingly more prevalent in modern neuroimaging studies. Despite numerous benefits from the growth in sample size, substantial technical variability associated with site/scanner-related effects exists which may inadvertently bias subsequent downstream analyses. Such a challenge calls for a data harmonization procedure which reduces the scanner effects and allows the scans to be combined for pooled analyses. In this work, we present MISPEL (Multi-scanner Image harmonization via Structure Preserving Embedding Learning), a multi-scanner harmonization framework. Unlike existing techniques, MISPEL does not assume a perfect coregistration across the scans, and the framework is naturally extendable to more than two scanners. Importantly, we incorporate our multi-scanner dataset where each subject is scanned on four different scanners. This unique paired dataset allows us to define and aim for an ideal harmonization (e.g., each subject with identical brain tissue volumes on all scanners). We extensively view scanner effects under varying metrics and demonstrate how MISPEL significantly improves them.

Realistic image normalization for multi-Domain segmentation

Image normalization is a building block in medical image analysis. Conventional approaches are customarily employed on a per-dataset basis. This strategy, however, prevents the current normalization algorithms from fully exploiting the complex joint information available across multiple datasets. Consequently, ignoring such joint information has a direct impact on the processing of segmentation algorithms. This paper proposes to revisit the conventional image normalization approach by, instead, learning a common normalizing function across multiple datasets. Jointly normalizing multiple datasets is shown to yield consistent normalized images as well as an improved image segmentation when intensity shifts are large. To do so, a fully automated adversarial and task-driven normalization approach is employed as it facilitates the training of realistic and interpretable images while keeping performance on par with the state-of-the-art. The adversarial training of our network aims at finding the optimal transfer function to improve both, jointly, the segmentation accuracy and the generation of realistic images. We have evaluated the performance of our normalizer on both infant and adult brain images from the iSEG, MRBrainS and ABIDE datasets. The results indicate that our contribution does provide an improved realism to the normalized images, while retaining a segmentation accuracy at par with the state-of-the-art learnable normalization approaches.

Quantification of Cervical Cord Cross-Sectional Area: Which Acquisition, Vertebra Level, and Analysis Software? A Multicenter Repeatability Study on a Traveling Healthy Volunteer

Background: Considerable spinal cord (SC) atrophy occurs in multiple sclerosis (MS). While MRI-based techniques for SC cross-sectional area (CSA) quantification have improved over time, there is no common agreement on whether to measure at single vertebral levels or across larger regions and whether upper SC CSA can be reliably measured from brain images. Aim: To compare in a multicenter setting three CSA measurement methods in terms of repeatability at different anatomical levels. To analyze the agreement between measurements performed on the cervical cord and on brain MRI. Method: One healthy volunteer was scanned three times on the same day in six sites (three scanner vendors) using a 3T MRI protocol including sagittal 3D T1-weighted imaging of the brain (covering the upper cervical cord) and of the SC. Images were analyzed using two semiautomated methods [NeuroQLab (NQL) and the Active Surface Model (ASM)] and the fully automated Spinal Cord Toolbox (SCT) on different vertebral levels (C1-C2; C2/3) on SC and brain images and the entire cervical cord (C1-C7) on SC images only. Results: CSA estimates were significantly smaller using SCT compared to NQL and ASM (p < 0.001), regardless of the cord level. Inter-scanner repeatability was best in C1-C7: coefficients of variation for NQL, ASM, and SCT: 0.4, 0.6, and 1.0%, respectively. CSAs estimated in brain MRI were slightly lower than in SC MRI (all p ≤ 0.006 at the C1-C2 level). Despite protocol harmonization between the centers with regard to image resolution and use of high-contrast 3D T1-weighted sequences, the variability of CSA was partly scanner dependent probably due to differences in scanner geometry, coil design, and details of the MRI parameter settings. Conclusion: For CSA quantification, dedicated isotropic SC MRI should be acquired, which yielded best repeatability in the entire cervical cord. In the upper part of the cervical cord, use of brain MRI scans entailed only a minor loss of CSA repeatability compared to SC MRI. Due to systematic differences between scanners and the CSA quantification software, both should be kept constant within a study. The MRI dataset of this study is available publicly to test new analysis approaches.

Voxel-Based Morphometry of Cerebellar Lobules in Essential Tremor

BACKGROUND

The extent of neurodegeneration underlying essential tremor (ET) remains a matter of debate. Despite various extents of cerebellar atrophy on structural magnetic resonance imaging (MRI), previous studies have shown substantial heterogeneity and included a limited number of patients. Novel automated pipelines allow detailed segmentation of cerebellar lobules based on structural MRI.

OBJECTIVE

To compare the volumes of cerebellar lobules in ET patients with those in healthy controls (HCs) using an automated segmentation pipeline.

METHODS

Structural MRI scans of ET patients eligible for deep brain stimulation (n = 55) and of age-matched and gender-matched HCs (n = 55, from the IXI database) were segmented using the automated CEREbellum Segmentation pipeline. Lobule-specific volume differences between the ET and HC groups were evaluated using a general linear model corrected for multiple tests.

RESULTS

Total brain tissue volumes did not differ between the ET and HC groups. ET patients demonstrated reduced volumes of lobules I-II, left Crus II, left VIIB, and an increased volume of right X when compared with the HC group.

CONCLUSION

A large cohort of ET patients demonstrated subtle signs of decreased cerebellar lobule volumes. These findings oppose the hypothesis of localized atrophy in cerebellar motor areas in ET, but not the possibility of cerebellar pathophysiology in ET. Prospective investigations using alternative neuroimaging modalities may further elucidate the pathophysiology of ET and provide insights into diagnostic and therapeutic approaches.

Bootstrap approach for meta-synthesis of MRI findings from multiple scanners

BACKGROUND

Neuroimaging data from large epidemiologic cohort studies often come from multiple scanners. The variations of MRI measurements due to differences in magnetic field strength, image acquisition protocols, and scanner vendors can influence the interpretation of aggregated data. The purpose of the present study was to compare methods that meta-analyze findings from a small number of different MRI scanners.

METHODS

We proposed a bootstrap resampling method using individual participant data and compared it with two common random effects meta-analysis methods, DerSimonian-Laird and Hartung-Knapp, and a conventional pooling method that combines MRI data from different scanners. We first performed simulations to compare the power and coverage probabilities of the four methods in the absence and presence of scanner effects on measurements. We then examined the association of age with white matter hyperintensity (WMH) volumes from 787 participants.

RESULTS

In simulations, the bootstrap approach performed better than the other three methods in terms of coverage probability and power when scanner differences were present. However, the bootstrap approach was consistent with pooling, the optimal approach, when scanner differences were absent. In the association of age with WMH volume, we observed that age was significantly associated with WMH volumes using the bootstrap approach, pooling, and the DerSimonian-Laird method, but not using the Hartung-Knapp method (p < 0.0001 for the bootstrap approach, DerSimonian-Laird, and pooling but p = 0.1439 for the Hartung-Knapp approach).

CONCLUSION

The bootstrap approach using individual participant data is suitable for integrating outcomes from multiple MRI scanners regardless of absence or presence of scanner effects on measurements.

Harmonized Segmentation of Neonatal Brain MRI

Deep learning based medical image segmentation has shown great potential in becoming a key part of the clinical analysis pipeline. However, many of these models rely on the assumption that the train and test data come from the same distribution. This means that such methods cannot guarantee high quality predictions when the source and target domains are dissimilar due to different acquisition protocols, or biases in patient cohorts. Recently, unsupervised domain adaptation techniques have shown great potential in alleviating this problem by minimizing the shift between the source and target distributions, without requiring the use of labeled data in the target domain. In this work, we aim to predict tissue segmentation maps on T 2-weighted magnetic resonance imaging data of an unseen preterm-born neonatal population, which has both different acquisition parameters and population bias when compared to our training data. We achieve this by investigating two unsupervised domain adaptation techniques with the objective of finding the best solution for our problem. We compare the two methods with a baseline fully-supervised segmentation network and report our results in terms of Dice scores obtained on our source test dataset. Moreover, we analyse tissue volumes and cortical thickness measures of the harmonized data on a subset of the population matched for gestational age at birth and postmenstrual age at scan. Finally, we demonstrate the applicability of the harmonized cortical gray matter maps with an analysis comparing term and preterm-born neonates and a proof-of-principle investigation of the association between cortical thickness and a language outcome measure.

Integrating large-scale neuroimaging research datasets: Harmonisation of white matter hyperintensity measurements across Whitehall and UK Biobank datasets

•

We harmonised measures of WMHs across two studies on healthy ageing.

•

Specific pre-processing strategies can increase comparability across studies.

•

Modelling of biological differences is crucial to provide calibrated measures.

Large scale neuroimaging datasets present the possibility of providing normative distributions for a wide variety of neuroimaging markers, which would vastly improve the clinical utility of these measures. However, a major challenge is our current poor ability to integrate measures across different large-scale datasets, due to inconsistencies in imaging and non-imaging measures across the different protocols and populations. Here we explore the harmonisation of white matter hyperintensity (WMH) measures across two major studies of healthy elderly populations, the Whitehall II imaging sub-study and the UK Biobank. We identify pre-processing strategies that maximise the consistency across datasets and utilise multivariate regression to characterise study sample differences contributing to differences in WMH variations across studies. We also present a parser to harmonise WMH-relevant non-imaging variables across the two datasets. We show that we can provide highly calibrated WMH measures from these datasets with: (1) the inclusion of a number of specific standardised processing steps; and (2) appropriate modelling of sample differences through the alignment of demographic, cognitive and physiological variables. These results open up a wide range of applications for the study of WMHs and other neuroimaging markers across extensive databases of clinical data.

MIL normalization -- prerequisites for accurate MRI radiomics analysis

The quality of magnetic resonance (MR) images obtained with different instruments and imaging parameters varies greatly. A large number of heterogeneous images are collected, and they suffer from acquisition variation. Such imaging quality differences will have a great impact on the radiomics analysis. The main differences in MR images include modality mismatch (M), intensity distribution variance (I), and layer-spacing differences (L), which are referred to as MIL differences in this paper for convenience. An MIL normalization system is proposed to reconstruct uneven MR images into high-quality data with complete modality, a uniform intensity distribution and consistent layer spacing. Three radiomics tasks, including tumor segmentation, pathological grading and genetic diagnosis of glioma, were used to verify the effect of MIL normalization on radiomics analysis. Three retrospective glioma datasets were analyzed in this study: BraTs (285 cases), TCGA (112 cases) and HuaShan (403 cases). They were used to test the effect of MIL on three different radiomics tasks, including tumor segmentation, pathological grading and genetic diagnosis. MIL normalization included three components: multimodal synthesis based on an encoder-decoder network, intensity normalization based on CycleGAN, and layer-spacing unification based on Statistical Parametric Mapping (SPM). The Dice similarity coefficient, areas under the curve (AUC) and six other indicators were calculated and compared after different normalization steps. The MIL normalization system can improved the Dice coefficient of segmentation by 9% (P < .001), the AUC of pathological grading by 32% (P < .001), and IDH1 status prediction by 25% (P < .001) when compared to non-normalization. The proposed MIL normalization system provides high-quality standardized data, which is a prerequisite for accurate radiomics analysis.

Clinical feasibility of longitudinal lateral ventricular volume measurements on T2-FLAIR across MRI scanner changes

•

Central and whole brain atrophy are faster in MS patients with disability progression.

•

These measures can be reliably assessed on clinically-available FLAIR images.

•

They are meaningful even with longitudinal scanner and field strength changes.

Background

Greater brain atrophy is associated with disability progression (DP) in patients with multiple sclerosis (PwMS). However, methodological challenges limit its routine clinical use.

Objective

To determine the feasibility of atrophy measures as markers of DP in PwMS scanned across different MRI field strengths.

Methods

A total of 980 PwMS were scanned on either 1.5 T or 3.0 T MRI scanners. Demographic and clinical data were retrospectively collected, and the presence of DP was determined according to standard clinical trial criteria. Lateral ventricular volume (LVV) change was measured with the NeuroSTREAM technique on clinical routine T2-FLAIR images. Percent brain volume change (PBVC) was measured using SIENA and ventricular cerebrospinal fluid (vCSF) % change was measured using VIENA and SIENAX algorithms on 3D T1-weighted images (WI). Stable vs. DP PwMS were compared using analysis of covariance (ANCOVA). Mixed modeling determined the effect of MRI scanner change on MRI-derived atrophy measures.

Results

Longitudinal LVV analysis was successful in all PwMS. SIENA-based PBVC and VIENA-based changes failed in 37.6% of cases, while SIENAX-based vCSF failed in 12.9% of cases. PwMS with DP (n = 241) had significantly greater absolute (20.9% vs. 8.7%, d = 0.66, p < 0.001) and annualized LVV % change (4.1% vs. 2.3%, d = 0.27, p < 0.001) when compared to stable PwMS (n = 739). In subjects with both analyses available, both 3D-T1 and T2-FLAIR-based analyses differentiated PwMS with DP (n = 149). However, only NeuroSTREAM and VIENA-based LVV/vCSF were able to show greater atrophy in PwMS that were scanned on different scanners. PBVC and SIENAX-based vCSF % changes were significantly affected by scanner change (Beta = −0.16, t-statistics = −2.133, p = 0.033 and Beta = −2.08, t-statistics = −4.084, p < 0.001), whereas no MRI scanner change effects on NeuroSTREAM-based PLVVC and VIENA-based vCSF % change were noted.

Conclusions

LVV-based atrophy on T2-FLAIR is a clinically relevant measure in spite of MRI scanner changes and mild disability levels.

Interpretation of Brain Volume Increase in Multiple Sclerosis

BACKGROUND AND PURPOSE

A high variability of brain MRI volume change measurement renders challenging its interpretation in multiple sclerosis (MS). Occurrence and clinical relevance of observed apparent brain volume increase (BVI) in MS patients have not been investigated yet. The objective was to quantify the prevalence and factors associated with BVI.

METHODS

We examined 366 MS patients (2,317 scans) and 44 controls (132 scans). Volumetric analysis of brain volume changes was performed by SIENA and ScanView. BVI was defined as brain volume change >0%. We compared characteristics of patients with and without BVI.

RESULTS

BVI was found in 26.3% (from 1,951) longitudinal scans (SIENA). If BVI occurred, a probability that BVI will be repeated consecutively more than or equal to two times was 15.9%. The repeated BVI was associated with clinical disease activity in 50% of cases. BVI was associated with shorter time and lower T2 lesion volume increase between two MRI scans, and higher normalized brain volume (all P < .0001). A proportion of scans with BVI was higher when analyzed by ScanView (35.3%) and in controls (36.4% by SIENA).

CONCLUSIONS

BVI occurs in a great proportion of MR scans over short-term follow-up and is not associated with disease stabilization. Although BVI can be caused by several factors, the results indicate that measurement error may contribute to BVI in the majority of cases. Clinicians should be aware of the frequent occurrence of apparent BVI, interpret brain volume changes in MS patients with great caution, and use methods with precise quantification of brain volume changes.

Voxel-Based Morphometry-from Hype to Hope. A Study on Hippocampal Atrophy in Mesial Temporal Lobe Epilepsy

BACKGROUND AND PURPOSE

Automated volumetry of the hippocampus is considered useful to assist the diagnosis of hippocampal sclerosis in temporal lobe epilepsy. However, voxel-based morphometry is rarely used for individual subjects because of high rates of false-positives. We investigated whether an approach with high dimensional warping to the template and nonparametric statistics would be useful to detect hippocampal atrophy in patients with hippocampal sclerosis.

MATERIALS AND METHODS

We performed single-subject voxel-based morphometry with nonparametric statistics within the framework of Statistical Parametric Mapping to compare MRI from 26 well-characterized patients with temporal lobe epilepsy individually against a group of 110 healthy controls. The following statistical threshold was used: P < .05 corrected for multiple comparisons with family-wise error over the region of interest right and left hippocampus.

RESULTS

The sensitivity for the detection of atrophy related to hippocampal sclerosis was 0.92 (95% CI, 0.67-0.99) for the right hippocampus and 0.60 (0.31-0.83) for the left, and the specificity for volume changes was 0.98 (0.93-0.99). All clusters of decreased hippocampal volumes were correctly lateralized to the seizure focus. Hippocampal volume decrease was in accordance with neuronal cell loss on histology reports.

CONCLUSIONS

Nonparametric voxel-based morphometry is sensitive and specific for hippocampal atrophy in patients with mesial temporal lobe epilepsy and may be useful in clinical practice.

Generalizing deep whole-brain segmentation for post-contrast MRI with transfer learning

Purpose: Generalizability is an important problem in deep neural networks, especially with variability of data acquisition in clinical magnetic resonance imaging (MRI). Recently, the spatially localized atlas network tiles (SLANT) can effectively segment whole brain, non-contrast T1w MRI with 132 volumetric labels. Transfer learning (TL) is a commonly used domain adaptation tool to update the neural network weights for local factors, yet risks degradation of performance on the original validation/test cohorts. Approach: We explore TL using unlabeled clinical data to address these concerns in the context of adapting SLANT to scanning protocol variations. We optimize whole-brain segmentation on heterogeneous clinical data by leveraging 480 unlabeled pairs of clinically acquired T1w MRI with and without intravenous contrast. We use labels generated on the pre-contrast image to train on the post-contrast image in a five-fold cross-validation framework. We further validated on a withheld test set of 29 paired scans over a different acquisition domain. Results: Using TL, we improve reproducibility across imaging pairs measured by the reproducibility Dice coefficient (rDSC) between the pre- and post-contrast image. We showed an increase over the original SLANT algorithm (rDSC 0.82 versus 0.72) and the FreeSurfer v6.0.1 segmentation pipeline ( rDSC = 0.53 ). We demonstrate the impact of this work decreasing the root-mean-squared error of volumetric estimates of the hippocampus between paired images of the same subject by 67%. Conclusion: This work demonstrates a pipeline for unlabeled clinical data to translate algorithms optimized for research data to generalize toward heterogeneous clinical acquisitions.

Intensity warping for multisite MRI harmonization

In multisite neuroimaging studies there is often unwanted technical variation across scanners and sites. These "scanner effects" can hinder detection of biological features of interest, produce inconsistent results, and lead to spurious associations. We propose mica (multisite image harmonization by cumulative distribution function alignment), a tool to harmonize images taken on different scanners by identifying and removing within-subject scanner effects. Our goals in the present study were to (1) establish a method that removes scanner effects by leveraging multiple scans collected on the same subject, and, building on this, (2) develop a technique to quantify scanner effects in large multisite studies so these can be reduced as a preprocessing step. We illustrate scanner effects in a brain MRI study in which the same subject was measured twice on seven scanners, and assess our method's performance in a second study in which ten subjects were scanned on two machines. We found that unharmonized images were highly variable across site and scanner type, and our method effectively removed this variability by aligning intensity distributions. We further studied the ability to predict image harmonization results for a scan taken on an existing subject at a new site using cross-validation.

Longitudinal ComBat: A method for harmonizing longitudinal multi-scanner imaging data

While aggregation of neuroimaging datasets from multiple sites and scanners can yield increased statistical power, it also presents challenges due to systematic scanner effects. This unwanted technical variability can introduce noise and bias into estimation of biological variability of interest. We propose a method for harmonizing longitudinal multi-scanner imaging data based on ComBat, a method originally developed for genomics and later adapted to cross-sectional neuroimaging data. Using longitudinal cortical thickness measurements from 663 participants in the Alzheimer's Disease Neuroimaging Initiative (ADNI) study, we demonstrate the presence of additive and multiplicative scanner effects in various brain regions. We compare estimates of the association between diagnosis and change in cortical thickness over time using three versions of the ADNI data: unharmonized data, data harmonized using cross-sectional ComBat, and data harmonized using longitudinal ComBat. In simulation studies, we show that longitudinal ComBat is more powerful for detecting longitudinal change than cross-sectional ComBat and controls the type I error rate better than unharmonized data with scanner included as a covariate. The proposed method would be useful for other types of longitudinal data requiring harmonization, such as genomic data, or neuroimaging studies of neurodevelopment, psychiatric disorders, or other neurological diseases.

Sex-Specific Differences in Life Span Brain Volumes in Multiple Sclerosis

BACKGROUND AND PURPOSE

Numerous sex-specific differences in multiple sclerosis (MS) susceptibility, disease manifestation, disability progression, inflammation, and neurodegeneration have been previously reported. Previous magnetic resonance imaging (MRI) studies have shown structural differences between female and male MS brain volumes. To determine sex-specific global and tissue-specific brain volume throughout the MS life span in a real-world large MRI database.

METHODS

A total of 2,199 MS patients (female/male ratio of 1,651/548) underwent structural MRI imaging on either a 1.5-T or 3-T scanner. Global and tissue-specific volumes of whole brain (WBV), white matter, and gray matter (GMV) were determined by utilizing Structural Image Evaluation using Normalisation of Atrophy Cross-sectional (SIENAX). Lateral ventricular volume (LVV) was determined with the Neurological Software Tool for REliable Atrophy Measurement (NeuroSTREAM). General linear models investigated sex and age interactions, and post hoc comparative sex analyses were performed.

RESULTS

Despite being age-matched with female MS patents, a greater proportion of male MS patients were diagnosed with progressive MS and had lower normalized WBV (P < .001), GMV (P < .001), and greater LVV (P < .001). In addition to significant stand-alone main effects, an interaction between sex and age had an additional effect on the LVV (F-statistics = 4.53, P = .033) and GMV (F-statistics = 4.59, P = .032). The sex and age interaction was retained in both models of LVV (F-statistics = 3.31, P = .069) and GMV (F-statistics = 6.1, P = .003) when disease subtype and disease-modifying treatment (DMT) were also included. Although male MS patients presented with significantly greater LVV and lower GMV during the early and midlife period when compared to their female counterparts (P < .001 for LVV and P < .019 for GMV), these differences were nullified in 60+ years old patients. Similar findings were seen within a subanalysis of MS patients that were not on any DMT at the time of enrollment.

CONCLUSION

There are sex-specific differences in the LVV and GMV over the MS life span.

Harmonization of large MRI datasets for the analysis of brain imaging patterns throughout the lifespan

As medical imaging enters its information era and presents rapidly increasing needs for big data analytics, robust pooling and harmonization of imaging data across diverse cohorts with varying acquisition protocols have become critical. We describe a comprehensive effort that merges and harmonizes a large-scale dataset of 10,477 structural brain MRI scans from participants without a known neurological or psychiatric disorder from 18 different studies that represent geographic diversity. We use this dataset and multi-atlas-based image processing methods to obtain a hierarchical partition of the brain from larger anatomical regions to individual cortical and deep structures and derive age trends of brain structure through the lifespan (3-96 years old). Critically, we present and validate a methodology for harmonizing this pooled dataset in the presence of nonlinear age trends. We provide a web-based visualization interface to generate and present the resulting age trends, enabling future studies of brain structure to compare their data with this reference of brain development and aging, and to examine deviations from ranges, potentially related to disease.

Brain MRI Predicts Worsening Multiple Sclerosis Disability over 5 Years in the SUMMIT Study

BACKGROUND AND PURPOSE

Brain MRI-derived lesions and atrophy are related to multiple sclerosis (MS) disability. In the Serially Unified Multicenter MS Investigation (SUMMIT), from Brigham and Women's Hospital (BWH) and University of California, San Francisco (UCSF), we assessed whether MRI methodologic heterogeneity may limit the ability to pool multisite data sets to assess 5-year clinical-MRI associations.

METHODS

Patients with relapsing-remitting (RR) MS (n = 100 from each site) underwent baseline brain MRI and baseline and 5-year clinical evaluations. Patients were matched on sex (74 women each), age, disease duration, and Expanded Disability Status Scale (EDSS) score. MRI was performed with differences between sites in both acquisition (field strength, voxel size, pulse sequences), and postprocessing pipeline to assess brain parenchymal fraction (BPF) and T2 lesion volume (T2LV).

RESULTS

The UCSF cohort showed higher correlation than the BWH cohort between T2LV and disease duration. UCSF showed a higher inverse correlation between BPF and age than BWH. UCSF showed a higher inverse correlation than BWH between BPF and 5-year EDSS score. Both cohorts showed inverse correlations between BPF and T2LV, with no between-site difference. The pooled but not individual cohort data showed a link between a lower baseline BPF and the subsequent 5-year worsening in disability in addition to other stronger relationships in the data.

CONCLUSIONS

MRI acquisition and processing differences may result in some degree of heterogeneity in assessing brain lesion and atrophy measures in patients with MS. Pooling of data across sites is beneficial to correct for potential biases in individual data sets.