Fig 6 - available via license: CC BY
Content may be subject to copyright.
White matter microstructure in schizophrenia, major depressive disorder, and 22q11.2 deletion syndrome. a White matter microstructural abnormalities are shown, by tract, based on the largest-ever diffusion MRI studies of these three disorders. In schizophrenia (SCZ), fractional anisotropy, a measure of white matter microstructure, is lower in almost all individual regions, and in the full skeleton. In major depressive disorder (MDD), a weak pattern of effects is observed, again with MDD patients showing on average lower FA across the full white matter skeleton, when compared to controls. In comparisons between 22q11.2 deletion syndrome (22q11DS) and matched controls, by contrast, the average FA along the full white matter skeleton does not show systematic differences; instead, while some regions do show on average lower FA in affected individuals compared with controls, several white matter regions show higher FA. b Relative to appropriately matched groups of healthy controls (HC), group differences in fractional anisotropy are shown for ENIGMA's studies of SCZ, MDD (both in adults), and 22q11.2 deletion syndrome. [Data adapted, with permission of the authors and publishers, from Kelly et al. 56 , van Velzen et al. 67 , and Villalón-Reina et al. 17 ; a key to the tract names appears in the original papers; some tracts (i.e. the hippocampal portion of the cingulum) were omitted from the 22q11DS analysis as they were not consistently in the field of view for some cohorts of the working group].

White matter microstructure in schizophrenia, major depressive disorder, and 22q11.2 deletion syndrome. a White matter microstructural abnormalities are shown, by tract, based on the largest-ever diffusion MRI studies of these three disorders. In schizophrenia (SCZ), fractional anisotropy, a measure of white matter microstructure, is lower in almost all individual regions, and in the full skeleton. In major depressive disorder (MDD), a weak pattern of effects is observed, again with MDD patients showing on average lower FA across the full white matter skeleton, when compared to controls. In comparisons between 22q11.2 deletion syndrome (22q11DS) and matched controls, by contrast, the average FA along the full white matter skeleton does not show systematic differences; instead, while some regions do show on average lower FA in affected individuals compared with controls, several white matter regions show higher FA. b Relative to appropriately matched groups of healthy controls (HC), group differences in fractional anisotropy are shown for ENIGMA's studies of SCZ, MDD (both in adults), and 22q11.2 deletion syndrome. [Data adapted, with permission of the authors and publishers, from Kelly et al. 56 , van Velzen et al. 67 , and Villalón-Reina et al. 17 ; a key to the tract names appears in the original papers; some tracts (i.e. the hippocampal portion of the cingulum) were omitted from the 22q11DS analysis as they were not consistently in the field of view for some cohorts of the working group].

Source publication
Article
Full-text available
Abstract This review summarizes the last decade of work by the ENIGMA (Enhancing NeuroImaging Genetics through Meta Analysis) Consortium, a global alliance of over 1400 scientists across 43 countries, studying the human brain in health and disease. Building on large-scale genetic studies that discovered the first robustly replicated genetic loci as...

Context in source publication

Context 1
... effect sizes varied by tract and included significant reductions in the anterior corona radiata (d = 0.40) and corpus callosum (d = 0.39, specifically its body (d = 0.39) and genu (d = 0.37)), effects were observed throughout the brain, with peak reductions observed for the entire WM skeleton (d = 0.42). Figure 6 shows these findings alongside data from two other disorders for which ENIGMA published largescale DTI analyses, MDD 67 , and 22q11DS 17 . Fig. 4 ENIGMA's large-scale studies of nine brain disorders. ...

Citations

... Here, we map local molecular attributes ("molecular vulnerability") and global network connectivity ("connectomic vulnerability") to case versus control cortical thickness abnormalities of thirteen different neurological, psychiatric, and neurodevelopmental diseases and disorders from the ENIGMA consortium 25 . We consistently find that disorder-specific cortical abnormality is shaped more by the local molecular fingerprints of brain regions than network embedding. ...
... MICA-MNI/ENIGMA 96 33 , obesity (N = 1223 participants, N = 2917 controls) 34 , schizotypy (N = 3004 participants) 35 , and Parkinson's disease (N = 2367 participants, N = 1183 controls) 36 . The ENIGMA (Enhancing Neuroimaging Genetics through Meta-Analysis) Consortium is a data-sharing initiative that relies on standardised processing and analysis pipelines, such that disorder maps are comparable 25 . Altogether, over 21,000 participants were scanned across the thirteen disorders, against almost 26,000 controls. ...
... Complete MEG acquisition protocols can be found in the HCP S1200 Release Manual. For each subject, we computed the power of the run at the vertex level across six different frequency bands: delta (2-4 Hz), theta (5-7 Hz), alpha (8)(9)(10)(11)(12), beta (15)(16)(17)(18)(19)(20)(21)(22)(23)(24)(25)(26)(27)(28)(29), low gamma , and high gamma , using the open-source software, Brainstorm 165 . Each power band was then parcellated into 68 cortical regions 41 . ...
Article
Full-text available
Numerous brain disorders demonstrate structural brain abnormalities, which are thought to arise from molecular perturbations or connectome miswiring. The unique and shared contributions of these molecular and connectomic vulnerabilities to brain disorders remain unknown, and has yet to be studied in a single multi-disorder framework. Using MRI morphometry from the ENIGMA consortium, we construct maps of cortical abnormalities for thirteen neurodevelopmental, neurological, and psychiatric disorders from N = 21,000 participants and N = 26,000 controls, collected using a harmonised processing protocol. We systematically compare cortical maps to multiple micro-architectural measures, including gene expression, neurotransmitter density, metabolism, and myelination (molecular vulnerability), as well as global connectomic measures including number of connections, centrality, and connection diversity (connectomic vulnerability). We find a relationship between molecular vulnerability and white-matter architecture that drives cortical disorder profiles. Local attributes, particularly neurotransmitter receptor profiles, constitute the best predictors of both disorder-specific cortical morphology and cross-disorder similarity. Finally, we find that cross-disorder abnormalities are consistently subtended by a small subset of network epicentres in bilateral sensory-motor, inferior temporal lobe, precuneus, and superior parietal cortex. Collectively, our results highlight how local molecular attributes and global connectivity jointly shape cross-disorder cortical abnormalities. Changes to structural and functional connectivity can give rise to neurodegeneration and neurodevelopmental diseases. Here the authors investigate molecular and connectomic patterns in 13 different neurological, psychiatric and neurodevelopmental diseases from the ENIGMA consortium.
... Large-scale data sharing initiatives offer opportunities to improve robustness by synthesizing multiple data sources (Thompson et al., 2020). However, in the behavioral sciences, differences in psychometric evaluation can confound the aggregation of data (Houtkoop et al., 2018;Towse et al., 2021). ...
Article
Objective: The variety of instruments used to assess posttraumatic stress disorder (PTSD) allows for flexibility, but also creates challenges for data synthesis. The objective of this work was to use a multisite mega analysis to derive quantitative recommendations for equating scores across measures of PTSD severity. Method: Empirical Bayes harmonization and linear models were used to describe and mitigate site and covariate effects. Quadratic models for converting scores across PTSD assessments were constructed using bootstrapping and tested on hold out data. Results: We aggregated 17 data sources and compiled an n = 5,634 sample of individuals who were assessed for PTSD symptoms. We confirmed our hypothesis that harmonization and covariate adjustments would significantly improve inference of scores across instruments. Harmonization significantly reduced cross-dataset variance (28%, p < .001), and models for converting scores across instruments were well fit (median R² = 0.985) with an average root mean squared error of 1.46 on sum scores. Conclusions: These methods allow PTSD symptom severity to be placed on multiple scales and offers interesting empirical perspectives on the role of harmonization in the behavioral sciences. (PsycInfo Database Record (c) 2022 APA, all rights reserved).
... [1][2][3] Embracing the "big data" era, the medical imaging community has widely adopted artificial intelligence (AI) for data analysis, from traditional statistical methods to machine learning (ML) models, which provides promise toward clinical translation. [4][5][6] Statistical tools such as univariate and multivariate prediction models are empowered to learn the associations between structural/functional variability and cognitive/psychiatric symptomatology in the human brain. Notably, advanced AI techniques have been successfully utilized in numerous clinical applications, such as computer-aided diagnosis, disease biomarker identification, and personalized disease risk quantification, which are bound to further revolutionize medical research and clinical practice. ...
... Here, we introduce several main variants that have been applied in the neuroimaging studies that will be discussed in the following sections. 5 ...
Preprint
Full-text available
Generative adversarial networks (GANs) are one powerful type of deep learning models that have been successfully utilized in numerous fields. They belong to a broader family called generative methods, which generate new data with a probabilistic model by learning sample distribution from real examples. In the clinical context, GANs have shown enhanced capabilities in capturing spatially complex, nonlinear, and potentially subtle disease effects compared to traditional generative methods. This review appraises the existing literature on the applications of GANs in imaging studies of various neurological conditions, including Alzheimer's disease, brain tumors, brain aging, and multiple sclerosis. We provide an intuitive explanation of various GAN methods for each application and further discuss the main challenges, open questions, and promising future directions of leveraging GANs in neuroimaging. We aim to bridge the gap between advanced deep learning methods and neurology research by highlighting how GANs can be leveraged to support clinical decision making and contribute to a better understanding of the structural and functional patterns of brain diseases.
... Researchers and clinicians can pursue this goal via collaborations in multi-center studies. Another option is to take part in global alliances like the ENIGMA (Enhancing NeuroImaging Genetics through Meta-Analysis) consortium, where over 50 diverse working groups participate in post-hoc data pooling and analyses (Thompson et al., 2020). These efforts accounted for the largest sMRI studies to date in several neurodevelopmental disorders (e.g. ...
Article
Full-text available
Structural magnetic resonance imaging (sMRI) offers immense potential for increasing our understanding of how anatomical brain development relates to clinical symptoms and functioning in neurodevelopmental disorders. Clinical developmental sMRI may help identify neurobiological risk factors or markers that may ultimately assist in diagnosis and treatment. However, researchers and clinicians aiming to conduct sMRI studies of neurodevelopmental disorders face several methodological challenges. This review offers hands-on guidelines for clinical developmental sMRI. First, we present brain morphometry metrics and review evidence on typical developmental trajectories throughout adolescence, together with atypical trajectories in selected neurodevelopmental disorders. Next, we discuss challenges and good scientific practices in study design, image acquisition and analysis, and recent options to implement quality control. Finally, we discuss choices related to statistical analysis and interpretation of results. We call for greater completeness and transparency in the reporting of methods to advance understanding of structural brain alterations in neurodevelopmental disorders.
... Such studies exemplify next-generation science: previous studies within the ENIGMA Consortium have resulted in important insights in the neurobiology of psychiatric conditions. 17 These discoveries reflect the advantages of largescale data analyses for testing the reproducibility and robustness of neuroimaging findings. 17 We expect the current project to provide similar insights, increasing our understanding of the development of psychopathology in youth at risk. ...
... 17 These discoveries reflect the advantages of largescale data analyses for testing the reproducibility and robustness of neuroimaging findings. 17 We expect the current project to provide similar insights, increasing our understanding of the development of psychopathology in youth at risk. In addition, by preregistering the study in advance of performing the analyses, we hope to contribute to a reduction of the potential publication bias in the field and to advance a more complete scientific record on this topic (cf. ...
... Conditions such as chronic pain, depression and anxiety disorders continue to resist treatment, and scientists are yet to find purely biological models for these prevalent conditions [1,2]. As such, search for technology-driven strategies to facilitate citizen science, and patient-centered and personalized care has begun [3][4][5]. ...
Chapter
Full-text available
The aim of this systematic scoping review was to gain a better understanding of research trends in digital mental health care. We focused on comorbid conditions: depression, anxiety, and pain–which continue to affect an estimated 20% of world population and require complex and continuous social and medical care provisions. We searched all randomized controlled trials on PubMed until May 2021 for any articles that used a form of information and communication technology (ICT) in relation to primary outcomes anxiety, pain, depression, or stress. From 1285 articles that satisfied the inclusion criteria, 890 were randomized trials with nearly 70% satisfactory outcomes. For depression and anxiety, the most frequently reported, were web-based, or mobile apps used for self-monitoring, and guided interventions. For pain, VR-based interventions or games were more prevalent, especially as tools for distraction, or as stimuli for mechanistic studies of pain or anxiety. We discuss gaps in knowledge and challenges that relate to the human factors in digital health applications, and underline the need for a practical and conceptual framework for capturing and reporting such variations.
... Imaging genetics is an integrated method that links genetic and epigenetic variations to neuroimaging measures and enables the assessment of quantitative genetic mechanisms . Currently, large cohort studies such as Enhancing Neuro Imaging Genetics through Meta-Analysis (ENIGMA) Consortium and UK Biobank, showing advantages for testing reproducibility and robustness of findings, provide an excellent platform to identify genetic underpinnings of imaging phenotypes (Hofer et al., 2020;Thompson et al., 2020). Imaging genetics has been applied in neuropsychiatry areas and is beginning to provide insight into epilepsy fields. ...
Article
Full-text available
Epilepsy is a neurological network disease with genetics playing a much greater role than was previously appreciated. Unfortunately, the relationship between genetic basis and imaging phenotype is by no means simple. Imaging genetics integrates multidimensional datasets within a unified framework, providing a unique opportunity to pursue a global vision for epilepsy. This review delineates the current knowledge of underlying genetic mechanisms for brain networks in different epilepsy syndromes, particularly from a neural developmental perspective. Further, endophenotypes and their potential value are discussed. Finally, we highlight current challenges and provide perspectives for the future development of imaging genetics in epilepsy.
... Machine learning models have shown great promise for precision diagnosis, treatment prediction, and a number of other clinical applications [4][5][6][7][8] . This has led to increasing interest in building systems where such models can aid human experts for accurate and efficient decision making in clinical settings [9][10][11] . However, there are some key challenges to achieving this goal [12][13][14] . ...
Preprint
Full-text available
Despite the great promise that machine learning has offered in many fields of medicine, it has also raised concerns about potential biases and poor generalization across genders, age distributions, races and ethnicities, hospitals, and data acquisition equipment and protocols. In the current study, and in the context of three brain diseases, we provide experimental data which support that when properly trained, machine learning models can generalize well across diverse conditions and do not suffer from biases. Specifically, by using multi-study magnetic resonance imaging consortia for diagnosing Alzheimer's disease, schizophrenia, and autism spectrum disorder, we find that, the accuracy of well-trained models is consistent across different subgroups pertaining to attributes such as gender, age, and racial groups, as also different clinical studies. We find that models that incorporate multi-source data from demographic, clinical, genetic factors and cognitive scores are also unbiased. These models have better predictive accuracy across subgroups than those trained only with structural measures in some cases but there are also situations when these additional features do not help.
... Retrospective harmonization uses advanced statistical approaches to harmonize data collected retrospectively, as is the case with some working groups within the Enhancing NeuroImaging Genetics through Meta-Analysis (ENIGMA) Consortium focused on clinical phenotyping (e.g., Esopenko et al., 2021). Of note, although these big data approaches have been fruitful and likely lead to a better understanding of disease states (see Thompson et al., 2020), even these large consortium studies lack specific recommendations and analytic strategies for the harmonization of neuropsychological outcome data collected across cultures. ...
Article
Objective: In this position article, we highlight the importance of considering cultural and linguistic variables that influence neuropsychological test performance and the possible moderating impact on our understanding of brain/behavior relationships. Increasingly, neuropsychologists are realizing that cultural and language differences between countries, regions, and ethnic groups influence neuropsychological outcomes, as test scores may not have the same interpretative meaning across cultures. Furthermore, attempts to apply the same norms across diverse populations without accounting for culture and language variations will result in detrimental ethical dilemmas, such as misdiagnosis of clinical conditions and inaccurate interpretations of research outcomes. Given the lack of normative data for ethnically and linguistically diverse communities, it is often challenging to merge data across diverse populations to investigate research questions of global significance. Methodological Considerations: We highlight some of the inherent challenges, limitations, and opportunities for efforts to harmonize cross-cultural neuropsychological data. We also explore some of the cultural factors that should be considered when attempting to harmonize cross-cultural neuropsychological data, sources of variance that should be accounted for in data analyses, and the need to identify evaluative criteria for interpreting data outcomes of cross-cultural harmonization approaches. Conclusion: In the future, it will be important to further solidify principles for aggregating data across diverse cultural and linguistic cohorts, validate whether assumptions are being satisfied regarding the relationship between neuropsychological measures and the brain and/or behavior of individuals from diverse cultural and linguistic backgrounds, as well as methods for evaluating relative successful validation for data harmonization efforts. (PsycInfo Database Record (c) 2022 APA, all rights reserved).
... To increase the amount of data and statistical power of analyses, research groups join together into consortia [11]. However, the need to protect patient data makes data sharing very challenging. ...
... Even when large consortia are established, they often only perform meta-analysis using traditional statistical methods, as opposed to joint megaanalysis using deep learning methods. A paradigmatic example of larges-scale meta-analysis is the ENIGMA Consortium [11]. ...
Preprint
Full-text available
The amount of biomedical data continues to grow rapidly. However, the ability to collect data from multiple sites for joint analysis remains challenging due to security, privacy, and regulatory concerns. We present a Secure Federated Learning architecture, MetisFL, which enables distributed training of neural networks over multiple data sources without sharing data. Each site trains the neural network over its private data for some time, then shares the neural network parameters (i.e., weights, gradients) with a Federation Controller, which in turn aggregates the local models, sends the resulting community model back to each site, and the process repeats. Our architecture provides strong security and privacy. First, sample data never leaves a site. Second, neural parameters are encrypted before transmission and the community model is computed under fully-homomorphic encryption. Finally, we use information-theoretic methods to limit information leakage from the neural model to prevent a curious site from performing membership attacks. We demonstrate this architecture in neuroimaging. Specifically, we investigate training neural models to classify Alzheimer's disease, and estimate Brain Age, from magnetic resonance imaging datasets distributed across multiple sites, including heterogeneous environments where sites have different amounts of data, statistical distributions, and computational capabilities.