Enhancing studies of the connectome in autism using the autism brain imaging data exchange II

Abnormal amygdala functional connectivity and deep learning classification in multifrequency bands in autism spectrum disorder: A multisite functional magnetic resonance imaging study

Previous studies have explored resting-state functional connectivity (rs-FC) of the amygdala in patients with autism spectrum disorder (ASD). However, it remains unclear whether there are frequency-specific FC alterations of the amygdala in ASD and whether FC in specific frequency bands can be used to distinguish patients with ASD from typical controls (TCs). Data from 306 patients with ASD and 314 age-matched and sex-matched TCs were collected from 28 sites in the Autism Brain Imaging Data Exchange database. The bilateral amygdala, defined as the seed regions, was used to perform seed-based FC analyses in the conventional, slow-5, and slow-4 frequency bands at each site. Image-based meta-analyses were used to obtain consistent brain regions across 28 sites in the three frequency bands. By combining generative adversarial networks and deep neural networks, a deep learning approach was applied to distinguish patients with ASD from TCs. The meta-analysis results showed frequency band specificity of FC in ASD, which was reflected in the slow-5 frequency band instead of the conventional and slow-4 frequency bands. The deep learning results showed that, compared with the conventional and slow-4 frequency bands, the slow-5 frequency band exhibited a higher accuracy of 74.73%, precision of 74.58%, recall of 75.05%, and area under the curve of 0.811 to distinguish patients with ASD from TCs. These findings may help us to understand the pathological mechanisms of ASD and provide preliminary guidance for the clinical diagnosis of ASD.

Applications of Deep Learning to Neurodevelopment in Pediatric Imaging: Achievements and Challenges

Deep learning has achieved remarkable progress, particularly in neuroimaging analysis. Deep learning applications have also been extended from adult to pediatric medical images, and thus, this paper aims to present a systematic review of this recent research. We first introduce the commonly used deep learning methods and architectures in neuroimaging, such as convolutional neural networks, auto-encoders, and generative adversarial networks. A non-exhaustive list of commonly used publicly available pediatric neuroimaging datasets and repositories are included, followed by a categorical review of recent works in pediatric MRI-based deep learning studies in the past five years. These works are categorized into recognizing neurodevelopmental disorders, identifying brain and tissue structures, estimating brain age/maturity, predicting neurodevelopment outcomes, and optimizing MRI brain imaging and analysis. Finally, we also discuss the recent achievements and challenges on these applications of deep learning to pediatric neuroimaging.

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

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

Preserved navigation abilities and spatio-temporal memory in individuals with autism spectrum disorder

Cerebellar abnormalities have been reported in autism spectrum disorder (ASD). Beyond its role in hallmark features of ASD, the cerebellum and its connectivity with forebrain structures also play a role in navigation. However, the current understanding of navigation abilities in ASD is equivocal, as is the impact of the disorder on the functional anatomy of the cerebellum. In the present study, we investigated the navigation behavior of a population of ASD and typically developing (TD) adults related to their brain anatomy as assessed by structural and functional MRI at rest. We used the Starmaze task, which permits assessing and distinguishing two complex navigation behaviors, one based on allocentric learning and the other on egocentric learning of a route with multiple decision points. Compared to TD controls, individuals with ASD showed similar exploration, learning, and strategy performance and preference. In addition, there was no difference in the structural or functional anatomy of the cerebellar circuits involved in navigation between the two groups. The findings of our work suggest that navigation abilities, spatio-temporal memory, and their underlying circuits are preserved in individuals with ASD.

Resting-state functional MRI in multicenter studies on multiple sclerosis: a report on raw data quality and functional connectivity features from the Italian Neuroimaging Network Initiative

The Italian Neuroimaging Network Initiative (INNI) is an expanding repository of brain MRI data from multiple sclerosis (MS) patients recruited at four Italian MRI research sites. We describe the raw data quality of resting-state functional MRI (RS-fMRI) time-series in INNI and the inter-site variability in functional connectivity (FC) features after unified automated data preprocessing. MRI datasets from 489 MS patients and 246 healthy control (HC) subjects were retrieved from the INNI database. Raw data quality metrics included temporal signal-to-noise ratio (tSNR), spatial smoothness (FWHM), framewise displacement (FD), and differential variation in signals (DVARS). Automated preprocessing integrated white-matter lesion segmentation (SAMSEG) into a standard fMRI pipeline (fMRIPrep). FC features were calculated on pre-processed data and harmonized between sites (Combat) prior to assessing general MS-related alterations. Across centers (both groups), median tSNR and FWHM ranged from 47 to 84 and from 2.0 to 2.5, and median FD and DVARS ranged from 0.08 to 0.24 and from 1.06 to 1.22. After preprocessing, only global FC-related features were significantly correlated with FD or DVARS. Across large-scale networks, age/sex/FD-adjusted and harmonized FC features exhibited both inter-site and site-specific inter-group effects. Significant general reductions were obtained for somatomotor and limbic networks in MS patients (vs. HC). The implemented procedures provide technical information on raw data quality and outcome of fully automated preprocessing that might serve as reference in future RS-fMRI studies within INNI. The unified pipeline introduced little bias across sites and appears suitable for multisite FC analyses on harmonized network estimates.

Alteration of resting-state network dynamics in autism spectrum disorder based on leading eigenvector dynamics analysis

Background

Neurobiological models to explain the vulnerability of autism spectrum disorders (ASDs) are scarce, and previous resting-state functional magnetic resonance imaging (rs-fMRI) studies mostly examined static functional connectivity (FC). Given that FC constantly evolves, it is critical to probe FC dynamic differences in ASD patients.

Methods

We characterized recurring phase-locking (PL) states during rest in 45 ASD patients and 47 age- and sex-matched healthy controls (HCs) using Leading Eigenvector Dynamics Analysis (LEiDA) and probed the organization of PL states across different fine grain sizes.

Results

Our results identified five different groups of discrete resting-state functional networks, which can be defined as recurrent PL state overtimes. Specifically, ASD patients showed an increased probability of three PL states, consisting of the visual network (VIS), frontoparietal control network (FPN), default mode network (DMN), and ventral attention network (VAN). Correspondingly, ASD patients also showed a decreased probability of two PL states, consisting of the subcortical network (SUB), somatomotor network (SMN), FPN, and VAN.

Conclusion

Our findings suggested that the temporal reorganization of brain discrete networks was closely linked to sensory to cognitive systems of the brain. Our study provides new insights into the dynamics of brain networks and contributes to a deeper understanding of the neurological mechanisms of ASD.

Gabrb3 is required for the functional integration of pyramidal neuron subtypes in the somatosensory cortex

Dysfunction of gamma-aminobutyric acid (GABA)ergic circuits is strongly associated with neurodevelopmental disorders. However, it is unclear how genetic predispositions impact circuit assembly. Using in vivo two-photon and widefield calcium imaging in developing mice, we show that Gabrb3, a gene strongly associated with autism spectrum disorder (ASD) and Angelman syndrome (AS), is enriched in contralaterally projecting pyramidal neurons and is required for inhibitory function. We report that Gabrb3 ablation leads to a developmental decrease in GABAergic synapses, increased local network synchrony, and long-lasting enhancement in functional connectivity of contralateral-but not ipsilateral-pyramidal neuron subtypes. In addition, Gabrb3 deletion leads to increased cortical response to tactile stimulation at neonatal stages. Using human transcriptomics and neuroimaging datasets from ASD subjects, we show that the spatial distribution of GABRB3 expression correlates with atypical connectivity in these subjects. Our studies reveal a requirement for Gabrb3 during the emergence of interhemispheric circuits for sensory processing.

The Link Between Autism and Sex-Related Neuroanatomy, and Associated Cognition and Gene Expression

OBJECTIVE

The male preponderance in prevalence of autism is among the most pronounced sex ratios across neurodevelopmental conditions. The authors sought to elucidate the relationship between autism and typical sex-differential neuroanatomy, cognition, and related gene expression.

METHODS

Using a novel deep learning framework trained to predict biological sex based on T₁-weighted structural brain images, the authors compared sex prediction model performance across neurotypical and autistic males and females. Multiple large-scale data sets comprising T₁-weighted MRI data were employed at four stages of the analysis pipeline: 1) pretraining, with the UK Biobank sample (>10,000 individuals); 2) transfer learning and validation, with the ABIDE data sets (1,412 individuals, 5-56 years of age); 3) test and discovery, with the EU-AIMS/AIMS-2-TRIALS LEAP data set (681 individuals, 6-30 years of age); and 4) specificity, with the NeuroIMAGE and ADHD200 data sets (887 individuals, 7-26 years of age).

RESULTS

Across both ABIDE and LEAP, features positively predictive of neurotypical males were on average significantly more predictive of autistic males (ABIDE: Cohen's d=0.48; LEAP: Cohen's d=1.34). Features positively predictive of neurotypical females were on average significantly less predictive of autistic females (ABIDE: Cohen's d=1.25; LEAP: Cohen's d=1.29). These differences in sex prediction accuracy in autism were not observed in individuals with ADHD. In autistic females, the male-shifted neurophenotype was further associated with poorer social sensitivity and emotional face processing while also associated with gene expression patterns of midgestational cell types.

CONCLUSIONS

The results demonstrate an increased resemblance in both autistic male and female individuals' neuroanatomy with male-characteristic patterns associated with typically sex-differential social cognitive features and related gene expression patterns. The findings hold promise for future research aimed at refining the quest for biological mechanisms underpinning the etiology of autism.

ConvNets for automatic detection of polyglutamine SCAs from brain MRIs: state of the art applications

Polyglutamine spinocerebellar ataxias (polyQ SCAs) are a group of neurodegenerative diseases, clinically and genetically heterogeneous, characterized by loss of balance and motor coordination due to dysfunction of the cerebellum and its connections. The diagnosis of each type of polyQ SCA, alongside with genetic tests, includes medical images analysis, and its automation may help specialists to distinguish between each type. Convolutional neural networks (ConvNets or CNNs) have been recently used for medical image processing, with outstanding results. In this work, we present the main clinical and imaging features of polyglutamine SCAs, and the basics of CNNs. Finally, we review studies that have used this approach to automatically process brain medical images and may be applied to SCAs detection. We conclude by discussing the possible limitations and opportunities of using ConvNets for SCAs diagnose in the future.

Genetic Heterogeneity Shapes Brain Connectivity in Psychiatry

BACKGROUND

Polygenicity and genetic heterogeneity pose great challenges for studying psychiatric conditions. Genetically informed approaches have been implemented in neuroimaging studies to address this issue. However, the effects on functional connectivity of rare and common genetic risks for psychiatric disorders are largely unknown. Our objectives were to estimate and compare the effect sizes on brain connectivity of psychiatric genomic risk factors with various levels of complexity: oligogenic copy number variants (CNVs), multigenic CNVs, and polygenic risk scores (PRSs) as well as idiopathic psychiatric conditions and traits.

METHODS

Resting-state functional magnetic resonance imaging data were processed using the same pipeline across 9 datasets. Twenty-nine connectome-wide association studies were performed to characterize the effects of 15 CNVs (1003 carriers), 7 PRSs, 4 idiopathic psychiatric conditions (1022 individuals with autism, schizophrenia, bipolar conditions, or attention-deficit/hyperactivity disorder), and 2 traits (31,424 unaffected control subjects).

RESULTS

Effect sizes on connectivity were largest for psychiatric CNVs (estimates: 0.2-0.65 z score), followed by psychiatric conditions (0.15-0.42), neuroticism and fluid intelligence (0.02-0.03), and PRSs (0.01-0.02). Effect sizes of CNVs on connectivity were correlated to their effects on cognition and risk for disease (r = 0.9, p = 5.93 × 10-6). However, effect sizes of CNVs adjusted for the number of genes significantly decreased from small oligogenic to large multigenic CNVs (r = -0.88, p = 8.78 × 10-6). PRSs had disproportionately low effect sizes on connectivity compared with CNVs conferring similar risk for disease.

CONCLUSIONS

Heterogeneity and polygenicity affect our ability to detect brain connectivity alterations underlying psychiatric manifestations.

Altered global modular organization of intrinsic functional connectivity in autism arises from atypical node-level processing

Autism spectrum disorder (ASD) is a neurodevelopmental condition characterized by restricted interests and repetitive behaviors as well as social-communication deficits. These traits are associated with atypicality of functional brain networks. Modular organization in the brain plays a crucial role in network stability and adaptability for neurodevelopment. Previous neuroimaging research demonstrates discrepancies in studies of functional brain modular organization in ASD. These discrepancies result from the examination of mixed age groups. Furthermore, recent findings suggest that while much attention has been given to deriving atlases and measuring the connections between nodes, within node information may also be crucial in determining altered modular organization in ASD compared with typical development (TD). However, altered modular organization originating from systematic nodal changes are yet to be explored in younger children with ASD. Here, we used graph-theoretical measures to fill this knowledge gap. To this end, we utilized multicenter resting-state fMRI data collected from 5 to 10-year-old children-34 ASD and 40 TD obtained from the Autism Brain Image Data Exchange (ABIDE) I and II. We demonstrate that alterations in topological roles and modular cohesiveness are the two key properties of brain regions anchored in default mode, sensorimotor, and salience networks, and primarily relate to social and sensory deficits in children with ASD. These results demonstrate that atypical global network organization in children with ASD arises from nodal role changes, and contribute to the growing body of literature suggesting that there is interesting information within nodes providing critical markers of functional brain networks in autistic children.

An individual-level weighted artificial neural network method to improve the systematic bias in BrainAGE analysis

BrainAGE is a commonly used machine learning technique to measure the accelerated/delayed development pattern of human brain structure/function with neuropsychiatric disorders. However, recent studies have shown a systematic bias ("regression toward mean" effect) in the BrainAGE method, which indicates that the prediction error is not uniformly distributed across Chronological Ages: for the older individuals, the Brain Ages would be under-estimated but would be over-estimated for the younger individuals. In the present study, we propose an individual-level weighted artificial neural network method and apply it to simulation datasets (containing 5000 simulated subjects) and a real dataset (containing 135 subjects). Results show that compared with traditional machine learning methods, the individual-level weighted strategy can significantly reduce the "regression toward mean" effect, while the prediction performance can achieve the comparable level with traditional machine learning methods. Further analysis indicates that the sigmoid active function for artificial neural network shows better performance than the relu active function. The present study provides a novel strategy to reduce the "regression toward mean" effect of BrainAGE analysis, which is helpful to improve accuracy in exploring the atypical brain structure/function development pattern of neuropsychiatric disorders.

CerebNet: A fast and reliable deep-learning pipeline for detailed cerebellum sub-segmentation

Quantifying the volume of the cerebellum and its lobes is of profound interest in various neurodegenerative and acquired diseases. Especially for the most common spinocerebellar ataxias (SCA), for which the first antisense oligonculeotide-base gene silencing trial has recently started, there is an urgent need for quantitative, sensitive imaging markers at pre-symptomatic stages for stratification and treatment assessment. This work introduces CerebNet, a fully automated, extensively validated, deep learning method for the lobular segmentation of the cerebellum, including the separation of gray and white matter. For training, validation, and testing, T1-weighted images from 30 participants were manually annotated into cerebellar lobules and vermal sub-segments, as well as cerebellar white matter. CerebNet combines FastSurferCNN, a UNet-based 2.5D segmentation network, with extensive data augmentation, e.g. realistic non-linear deformations to increase the anatomical variety, eliminating additional preprocessing steps, such as spatial normalization or bias field correction. CerebNet demonstrates a high accuracy (on average 0.87 Dice and 1.742mm Robust Hausdorff Distance across all structures) outperforming state-of-the-art approaches. Furthermore, it shows high test-retest reliability (average ICC >0.97 on OASIS and Kirby) as well as high sensitivity to disease effects, including the pre-ataxic stage of spinocerebellar ataxia type 3 (SCA3). CerebNet is compatible with FreeSurfer and FastSurfer and can analyze a 3D volume within seconds on a consumer GPU in an end-to-end fashion, thus providing an efficient and validated solution for assessing cerebellum sub-structure volumes. We make CerebNet available as source-code (https://github.com/Deep-MI/FastSurfer).

Systematic evaluation of machine learning algorithms for neuroanatomically-based age prediction in youth

Application of machine learning (ML) algorithms to structural magnetic resonance imaging (sMRI) data has yielded behaviorally meaningful estimates of the biological age of the brain (brain-age). The choice of the ML approach in estimating brain-age in youth is important because age-related brain changes in this age-group are dynamic. However, the comparative performance of the available ML algorithms has not been systematically appraised. To address this gap, the present study evaluated the accuracy (mean absolute error [MAE]) and computational efficiency of 21 machine learning algorithms using sMRI data from 2105 typically developing individuals aged 5-22 years from five cohorts. The trained models were then tested in two independent holdout datasets, one comprising 4078 individuals aged 9-10 years and another comprising 594 individuals aged 5-21 years. The algorithms encompassed parametric and nonparametric, Bayesian, linear and nonlinear, tree-based, and kernel-based models. Sensitivity analyses were performed for parcellation scheme, number of neuroimaging input features, number of cross-validation folds, number of extreme outliers, and sample size. Tree-based models and algorithms with a nonlinear kernel performed comparably well, with the latter being especially computationally efficient. Extreme Gradient Boosting (MAE of 1.49 years), Random Forest Regression (MAE of 1.58 years), and Support Vector Regression (SVR) with Radial Basis Function (RBF) Kernel (MAE of 1.64 years) emerged as the three most accurate models. Linear algorithms, with the exception of Elastic Net Regression, performed poorly. Findings of the present study could be used as a guide for optimizing methodology when quantifying brain-age in youth.

Attenuated effective connectivity of large-scale brain networks in children with autism spectrum disorders

INTRODUCTION

Understanding the neurological basis of autism spectrum disorder (ASD) is important for the diagnosis and treatment of this mental disorder. Emerging evidence has suggested aberrant functional connectivity of large-scale brain networks in individuals with ASD. However, whether the effective connectivity which measures the causal interactions of these networks is also impaired in these patients remains unclear.

OBJECTS

The main purpose of this study was to investigate the effective connectivity of large-scale brain networks in patients with ASD during resting state.

MATERIALS AND METHODS

The subjects were 42 autistic children and 127 age-matched normal children from the ABIDE II dataset. We investigated effective connectivity of 7 large-scale brain networks including visual network (VN), default mode network (DMN), cerebellum, sensorimotor network (SMN), auditory network (AN), salience network (SN), frontoparietal network (FPN), with spectral dynamic causality model (spDCM). Parametric empirical Bayesian (PEB) was used to perform second-level group analysis and furnished group commonalities and differences in effective connectivity. Furthermore, we analyzed the correlation between the strength of effective connectivity and patients' clinical characteristics.

RESULTS

For both groups, SMN acted like a hub network which demonstrated dense effective connectivity with other large-scale brain network. We also observed significant causal interactions within the "triple networks" system, including DMN, SN and FPN. Compared with healthy controls, children with ASD showed decreased effective connectivity among some large-scale brain networks. These brain networks included VN, DMN, cerebellum, SMN, and FPN. In addition, we also found significant negative correlation between the strength of the effective connectivity from right angular gyrus (ANG_R) of DMN to left precentral gyrus (PreCG_L) of SMN and ADOS-G or ADOS-2 module 4 stereotyped behaviors and restricted interest total (ADOS_G_STEREO_BEHAV) scores.

CONCLUSION

Our research provides new evidence for the pathogenesis of children with ASD from the perspective of effective connections within and between large-scale brain networks. The attenuated effective connectivity of brain networks may be a clinical neurobiological feature of ASD. Changes in effective connectivity of brain network in children with ASD may provide useful information for the diagnosis and treatment of the disease.

Dysregulated dynamic time-varying triple-network segregation in children with autism spectrum disorder

One of the remarkable characteristics of autism spectrum disorder (ASD) is the dysregulation of functional connectivity of the triple-network, which includes the salience network (SN), default mode network (DMN), and central executive network (CEN). However, there is little known about the segregation of the triple-network dynamics in ASD. This study used resting-state functional magnetic resonance imaging data including 105 ASD and 102 demographically-matched typical developing control (TC) children. We compared the dynamic time-varying triple-network segregation and triple-network functional connectivity states between ASD and TC groups, and examined the relationship between dynamic triple-network segregation alterations and clinical symptoms of ASD. The average dynamic network segregation value of the DMN with SN and the DMN with CEN in ASD was lower but the coefficient of variation (CV) of dynamic network segregation of the DMN with CEN was higher in ASD. Furthermore, partially reduced triple-network segregation associated with the DMN was found in connectivity states analysis of ASD. These abnormal average values and CV of dynamic network segregation predicted social communication deficits and restricted and repetitive behaviors in ASD. Our findings indicate abnormal dynamic time-varying triple-network segregation of ASD and highlight the crucial role of the triple-network in the neural mechanisms underlying ASD.

Cofluctuation analysis reveals aberrant default mode network patterns in adolescents and youths with autism spectrum disorder

Resting-state functional connectivity (rsFC) approaches provide informative estimates of the functional architecture of the brain, and recently-proposed cofluctuation analysis temporally unwraps FC at every moment in time, providing refined information for quantifying brain dynamics. As a brain network disorder, autism spectrum disorder (ASD) was characterized by substantial alteration in FC, but the contribution of moment-to-moment-activity cofluctuations to the overall dysfunctional connectivity pattern in ASD remains poorly understood. Here, we used the cofluctuation approach to explore the underlying dynamic properties of FC in ASD, using a large multisite resting-state functional magnetic resonance imaging (rs-fMRI) dataset (ASD = 354, typically developing controls [TD] = 446). Our results verified that the networks estimated using high-amplitude frames were highly correlated with the traditional rsFC. Moreover, these frames showed higher average amplitudes in participants with ASD than those in the TD group. Principal component analysis was performed on the activity patterns in these frames and aggregated over all subjects. The first principal component (PC1) corresponds to the default mode network (DMN), and the PC1 coefficients were greater in participants with ASD than those in the TD group. Additionally, increased ASD symptom severity was associated with the increased coefficients, which may result in excessive internally oriented cognition and social cognition deficits in individuals with ASD. Our finding highlights the utility of cofluctuation approaches in prevalent neurodevelopmental disorders and verifies that the aberrant contribution of DMN to rsFC may underline the symptomatology in adolescents and youths with ASD.

Functional Connectome-Based Predictive Modeling in Autism

Autism is a heterogeneous neurodevelopmental condition, and functional magnetic resonance imaging-based studies have helped advance our understanding of its effects on brain network activity. We review how predictive modeling, using measures of functional connectivity and symptoms, has helped reveal key insights into this condition. We discuss how different prediction frameworks can further our understanding of the brain-based features that underlie complex autism symptomatology and consider how predictive models may be used in clinical settings. Throughout, we highlight aspects of study interpretation, such as data decay and sampling biases, that require consideration within the context of this condition. We close by suggesting exciting future directions for predictive modeling in autism.

Robust, Generalizable, and Interpretable Artificial Intelligence-Derived Brain Fingerprints of Autism and Social Communication Symptom Severity

BACKGROUND

Autism spectrum disorder (ASD) is among the most pervasive neurodevelopmental disorders, yet the neurobiology of ASD is still poorly understood because inconsistent findings from underpowered individual studies preclude the identification of robust and interpretable neurobiological markers and predictors of clinical symptoms.

METHODS

We leverage multiple brain imaging cohorts and exciting recent advances in explainable artificial intelligence to develop a novel spatiotemporal deep neural network (stDNN) model, which identifies robust and interpretable dynamic brain markers that distinguish ASD from neurotypical control subjects and predict clinical symptom severity.

RESULTS

stDNN achieved consistently high classification accuracies in cross-validation analysis of data from the multisite ABIDE (Autism Brain Imaging Data Exchange) cohort (n = 834). Crucially, stDNN also accurately classified data from independent Stanford (n = 202) and GENDAAR (Gender Exploration of Neurogenetics and Development to Advanced Autism Research) (n = 90) cohorts without additional training. stDNN could not distinguish attention-deficit/hyperactivity disorder from neurotypical control subjects, highlighting the model's specificity. Explainable artificial intelligence revealed that brain features associated with the posterior cingulate cortex and precuneus, dorsolateral and ventrolateral prefrontal cortex, and superior temporal sulcus, which anchor the default mode network, cognitive control, and human voice processing systems, respectively, most clearly distinguished ASD from neurotypical control subjects in the three cohorts. Furthermore, features associated with the posterior cingulate cortex and precuneus nodes of the default mode network emerged as robust predictors of the severity of core social and communication deficits but not restricted/repetitive behaviors in ASD.

CONCLUSIONS

Our findings, replicated across independent cohorts, reveal robust individualized functional brain fingerprints of ASD psychopathology, which could lead to more objective and precise phenotypic characterization and targeted treatments.

Machine learning for autism spectrum disorder diagnosis using structural magnetic resonance imaging: Promising but challenging

Autism spectrum disorder (ASD) is a complex neurodevelopmental disorder that affects approximately 1% of the population and causes significant burdens. ASD's pathogenesis remains elusive; hence, diagnosis is based on a constellation of behaviors. Structural magnetic resonance imaging (sMRI) studies have shown several abnormalities in volumetric and geometric features of the autistic brain. However, inconsistent findings prevented most contributions from being translated into clinical practice. Establishing reliable biomarkers for ASD using sMRI is crucial for the correct diagnosis and treatment. In recent years, machine learning (ML) and specifically deep learning (DL) have quickly extended to almost every sector, notably in disease diagnosis. Thus, this has led to a shift and improvement in ASD diagnostic methods, fulfilling most clinical diagnostic requirements. However, ASD discovery remains difficult. This review examines the ML-based ASD diagnosis literature over the past 5 years. A literature-based taxonomy of the research landscape has been mapped, and the major aspects of this topic have been covered. First, we provide an overview of ML's general classification pipeline and the features of sMRI. Next, representative studies are highlighted and discussed in detail with respect to methods, and biomarkers. Finally, we highlight many common challenges and make recommendations for future directions. In short, the limited sample size was the main obstacle; Thus, comprehensive data sets and rigorous methods are necessary to check the generalizability of the results. ML technologies are expected to advance significantly in the coming years, contributing to the diagnosis of ASD and helping clinicians soon.

The atypical functional connectivity of Broca's area at multiple frequency bands in autism spectrum disorder

As a developmental disorder, autism spectrum disorder (ASD) has drawn much attention due to its severe impacts on one's language capacity. Broca's area, an important brain region of the language network, is largely involved in language-related functions. Using the Autism Brain Image Data Exchange (ABIDE) dataset, a mega-analysis was performed involving a total of 1454 participants (including 618 individuals with ASD and 836 healthy controls (HCs). To detect the neural pathophysiological mechanism of ASD from the perspective of language, we conducted a functional connectivity (FC) analysis with Broca's area as the seed in multiple frequency bands (conventional: 0.01-0.08 Hz; slow-4: 0.027-0.073 Hz; slow-5: 0.01-0.027 Hz). We found that compared with HC, ASD patients demonstrated increased FC in the left thalamus, left precuneus, left anterior cingulate and paracingulate gyri, and left medial orbital of the superior frontal gyrus in the conventional frequency band (0.01-0.08 Hz). The results of the slow-5 frequency band (0.01-0.027 Hz) presented increased FC values of the left precuneus, left medial orbital of the superior frontal gyrus, right medial orbital of the superior frontal gyrus and right thalamus. No significant cluster was detected in the slow-4 frequency band (0.027-0.073 Hz). In conclusion, the abnormal functional connectivity in patients with ASD has frequency-specific properties. Furthermore, the slow-5 frequency band (0.01-0.027 Hz) mainly contributed to the findings of the conventional frequency band (0.01-0.08 Hz). The current study might shed new light on the neural pathophysiological mechanism of language impairments in people with ASD.

An Open MRI Dataset For Multiscale Neuroscience

Multimodal neuroimaging grants a powerful window into the structure and function of the human brain at multiple scales. Recent methodological and conceptual advances have enabled investigations of the interplay between large-scale spatial trends (also referred to as gradients) in brain microstructure and connectivity, offering an integrative framework to study multiscale brain organization. Here, we share a multimodal MRI dataset for Microstructure-Informed Connectomics (MICA-MICs) acquired in 50 healthy adults (23 women; 29.54 ± 5.62 years) who underwent high-resolution T1-weighted MRI, myelin-sensitive quantitative T1 relaxometry, diffusion-weighted MRI, and resting-state functional MRI at 3 Tesla. In addition to raw anonymized MRI data, this release includes brain-wide connectomes derived from (i) resting-state functional imaging, (ii) diffusion tractography, (iii) microstructure covariance analysis, and (iv) geodesic cortical distance, gathered across multiple parcellation scales. Alongside, we share large-scale gradients estimated from each modality and parcellation scale. Our dataset will facilitate future research examining the coupling between brain microstructure, connectivity, and function. MICA-MICs is available on the Canadian Open Neuroscience Platform data portal ( https://portal.conp.ca ) and the Open Science Framework ( https://osf.io/j532r/ ).

Transdiagnostic connectome signatures from resting-state fMRI predict individual-level intellectual capacity

Medication and other therapies for psychiatric disorders show unsatisfying efficacy, in part due to the significant clinical/ biological heterogeneity within each disorder and our over-reliance on categorical clinical diagnoses. Alternatively, dimensional transdiagnostic studies have provided a promising pathway toward realizing personalized medicine and improved treatment outcomes. One factor that may influence response to psychiatric treatments is cognitive function, which is reflected in one's intellectual capacity. Intellectual capacity is also reflected in the organization and structure of intrinsic brain networks. Using a large transdiagnostic cohort (n = 1721), we sought to discover neuroimaging biomarkers by developing a resting-state functional connectome-based prediction model for a key intellectual capacity measure, Full-Scale Intelligence Quotient (FSIQ), across the diagnostic spectrum. Our cross-validated model yielded an excellent prediction accuracy (r = 0.5573, p < 0.001). The robustness and generalizability of our model was further validated on three independent cohorts (n = 2641). We identified key transdiagnostic connectome signatures underlying FSIQ capacity involving the dorsal-attention, frontoparietal and default-mode networks. Meanwhile, diagnosis groups showed disorder-specific biomarker patterns. Our findings advance the neurobiological understanding of cognitive functioning across traditional diagnostic categories and provide a new avenue for neuropathological classification of psychiatric disorders.

Altered resting state dynamic functional connectivity of amygdala subregions in patients with autism spectrum disorder: A multi-site fMRI study

BACKGROUND

Autism spectrum disorder (ASD) is associated with altered brain connectivity. Previous studies have focused on the static functional connectivity pattern from amygdala subregions in ASD while ignoring its dynamics. Considering that dynamic functional connectivity (dFC) can provide different perspectives, the present study aims to investigate the dFC pattern of the amygdala subregions in ASD patients.

METHODS

Data of 618 ASD patients and 836 typical controls (TCs) of 30 sites were obtained from the Autism Brain Imaging Data Exchange (ABIDE) database. The sliding window approach was applied to conduct seed-based dFC analysis. The seed regions were bilateral basolateral (BLA) and centromedial-superficial amygdala (CSA). A two-sample t-test was done at each site. Image-based meta-analysis (IBMA) based on the results from all sites was performed. Correlation analysis was conducted between the dFC values and the clinical scores.

RESULTS

The ASD patients showed lower dFC between the left BLA and the bilateral inferior temporal (ITG)/left superior frontal gyrus, between the right BLA and right ITG/right thalamus/left superior temporal gyrus, and between the right CSA and middle temporal gyrus. The ASD patients showed higher dFC between the left BLA and temporal lobe/right supramarginal gyrus, between the right BLA and left calcarine gyrus, and between the left CSA and left calcarine gyrus. Correlation analysis revealed that the symptom severity was positively correlated with the dFC between the bilateral BLA and ITG in ASD.

CONCLUSIONS

Abnormal dFC of the specific amygdala subregions may provide new insights into the pathological mechanisms of ASD.

Examining Associations Between Amygdala Volumes and Anxiety Symptoms in Autism Spectrum Disorder

BACKGROUND

Anxiety is one of the most common co-occurring conditions in people with autism spectrum disorder. The amygdala has been identified as being associated with anxiety in populations with and without autism, yet associations in autism were based on relatively small or developmentally constrained samples, leaving questions as to whether these results hold at different developmental ages and in a larger, more robust sample.

METHODS

Structural neuroimaging and parent report of anxiety symptoms of children ages 5-13 years with (n = 123) and without (n = 171) a diagnosis of autism were collected from the University of Maryland and three sites from the Autism Brain Imaging Data Exchange. Standardized residuals for bilateral amygdala volumes were computed adjusting for site, hemispheric volumes, and covariates (age, sex, Full Scale IQ).

RESULTS

Clinically significant anxiety symptoms did not differentiate amygdala volumes between groups (i.e., autism and anxiety, autism without anxiety, without autism or anxiety). No significant association between left or right amygdala volumes and anxiety scores was observed among the sample of individuals with autism. Meta-analytic and Bayes factor estimations provided additional support for the null hypothesis. Age, sex, and autism severity did not moderate associations between anxiety and amygdala volumes.

CONCLUSIONS

No relation between amygdala volumes and anxiety symptoms in children with autism was observed in the largest sample to investigate this question. We discuss directions for future research to determine whether additional factors including age or method of assessment may contribute to this lack of association.

Brain volume abnormalities and clinical outcomes following paediatric traumatic brain injury

Long-term outcomes are difficult to predict after paediatric traumatic brain injury. The presence or absence of focal brain injuries often do not explain cognitive, emotional and behavioural disabilities that are common and disabling. In adults, traumatic brain injury produces progressive brain atrophy that can be accurately measured and is associated with cognitive decline. However, the effect of paediatric traumatic brain injury on brain volumes is more challenging to measure because of its interaction with normal brain development. Here we report a robust approach to the individualized estimation of brain volume following paediatric traumatic brain injury and investigate its relationship to clinical outcomes. We first used a large healthy control dataset (n > 1200, age 8-22) to describe the healthy development of white and grey matter regions through adolescence. Individual estimates of grey and white matter regional volume were then generated for a group of moderate/severe traumatic brain injury patients injured in childhood (n = 39, mean age 13.53 ± 1.76, median time since injury = 14 months, range 4-168 months) by comparing brain volumes in patients to age-matched controls. Patients were individually classified as having low or normal brain volume. Neuropsychological and neuropsychiatric outcomes were assessed using standardized testing and parent/carer assessments. Relative to head size, grey matter regions decreased in volume during normal adolescence development whereas white matter tracts increased in volume. Traumatic brain injury disrupted healthy brain development, producing reductions in both grey and white matter brain volumes after correcting for age. Of the 39 patients investigated, 11 (28%) had at least one white matter tract with reduced volume and seven (18%) at least one area of grey matter with reduced volume. Those classified as having low brain volume had slower processing speed compared to healthy controls, emotional impairments, higher levels of apathy, increased anger and learning difficulties. In contrast, the presence of focal brain injury and microbleeds were not associated with an increased risk of these clinical impairments. In summary, we show how brain volume abnormalities after paediatric traumatic brain injury can be robustly calculated from individual T1 MRI using a large normative dataset that allows the effects of healthy brain development to be controlled for. Using this approach, we show that volumetric abnormalities are common after moderate/severe traumatic brain injury in both grey and white matter regions, and are associated with higher levels of cognitive, emotional and behavioural abnormalities that are common after paediatric traumatic brain injury.

Accounting for motion in resting-state fMRI: What part of the spectrum are we characterizing in autism spectrum disorder?

The exclusion of high-motion participants can reduce the impact of motion in functional Magnetic Resonance Imaging (fMRI) data. However, the exclusion of high-motion participants may change the distribution of clinically relevant variables in the study sample, and the resulting sample may not be representative of the population. Our goals are two-fold: 1) to document the biases introduced by common motion exclusion practices in functional connectivity research and 2) to introduce a framework to address these biases by treating excluded scans as a missing data problem. We use a study of autism spectrum disorder in children without an intellectual disability to illustrate the problem and the potential solution. We aggregated data from 545 children (8-13 years old) who participated in resting-state fMRI studies at Kennedy Krieger Institute (173 autistic and 372 typically developing) between 2007 and 2020. We found that autistic children were more likely to be excluded than typically developing children, with 28.5% and 16.1% of autistic and typically developing children excluded, respectively, using a lenient criterion and 81.0% and 60.1% with a stricter criterion. The resulting sample of autistic children with usable data tended to be older, have milder social deficits, better motor control, and higher intellectual ability than the original sample. These measures were also related to functional connectivity strength among children with usable data. This suggests that the generalizability of previous studies reporting naïve analyses (i.e., based only on participants with usable data) may be limited by the selection of older children with less severe clinical profiles because these children are better able to remain still during an rs-fMRI scan. We adapt doubly robust targeted minimum loss based estimation with an ensemble of machine learning algorithms to address these data losses and the resulting biases. The proposed approach selects more edges that differ in functional connectivity between autistic and typically developing children than the naïve approach, supporting this as a promising solution to improve the study of heterogeneous populations in which motion is common.

Diagnosis-informed connectivity subtyping discovers subgroups of autism with reproducible symptom profiles

Clinical heterogeneity has been one of the main barriers to develop effective biomarkers and therapeutic strategies in autism spectrum disorder (ASD). Recognizing this challenge, much effort has been made in recent neuroimaging studies to find biologically more homogeneous subgroups (called 'neurosubtypes') in autism. However, most approaches have rarely evaluated how much the employed features in subtyping represent the core anomalies of ASD, obscuring its utility in actual clinical diagnosis. To address this, we combined two data-driven methods, 'connectome-based gradient' and 'functional random forest', collectively allowing to discover reproducible neurosubtypes based on resting-state functional connectivity profiles that are specific to ASD. Indeed, the former technique provides the features (as input for subtyping) that effectively summarize whole-brain connectome variations in both normal and ASD conditions, while the latter leverages a supervised random forest algorithm to inform diagnostic labels to clustering, which makes neurosubtyping driven by the features of ASD core anomalies. Applying this framework to the open-sharing Autism Brain Imaging Data Exchange repository data (discovery, n = 103/108 for ASD/typically developing [TD]; replication, n = 44/42 for ASD/TD), we found three dominant subtypes of functional gradients in ASD and three subtypes in TD. The subtypes in ASD revealed distinct connectome profiles in multiple brain areas, which are associated with different Neurosynth-derived cognitive functions previously implicated in autism studies. Moreover, these subtypes showed different symptom severity, which degree co-varies with the extent of functional gradient changes observed across the groups. The subtypes in the discovery and replication datasets showed similar symptom profiles in social interaction and communication domains, confirming a largely reproducible brain-behavior relationship. Finally, the connectome gradients in ASD subtypes present both common and distinct patterns compared to those in TD, reflecting their potential overlap and divergence in terms of developmental mechanisms involved in the manifestation of large-scale functional networks. Our study demonstrated a potential of the diagnosis-informed subtyping approach in developing a clinically useful brain-based classification system for future ASD research.

Systematic Bibliometric and Visualized Analysis of Research Hotspots and Trends on Autism Spectrum Disorder Neuroimaging

Background

Autism spectrum disorder (ASD) is a chronic developmental disability caused by differences in the brain. The gold standard for the diagnosis of this condition is based on behavioral science, but research on the application of neurological detection to diagnose the atypical nervous system of ASD is ongoing. ASD neuroimaging research involves the examination of the brain's structure, functional connections, and neurometabolic. However, limited medical resource and the unique heterogeneity of ASD have resulted in many challenges when neuroimaging is utilized.

Objective

This bibliometric study is aimed at summarizing themes and trends in research on autism spectrum disorder neuroimaging and at proposing potential directions for future inquiry.

Methods

Citations were downloaded from the Web of Science Core Collection database on neuroimaging published from January 1, 2012, to December 31, 2021. The retrieved information was analyzed using Bibliometric.com, CiteSpace.5.8. R3, and VOS viewer.

Results

A total of 1,363 papers were published across 58 regions. The United States was the leading source of publications. The League of European Research Universities published the largest number of articles (171). Burst keywords from 2018 to 2021 include identification and network. The clusters of references that continued into 2020 included graph theory, functional connectivity, and classification, which represent key research topics.

Conclusions

Imaging data is being used to identify neuro-network models with higher accuracy for ASD discrimination. Functional near-infrared imaging is advantageous compared to other neuroimaging. In the future, research on systematic and accurate computer-aided diagnosis technology should be encouraged. Moreover, the study of neuroimaging of ASD in different psychological and behavioral states can inspire new ideas about the diagnosis and intervention training of ASD and should be explored.

The Atypical Effective Connectivity of Right Temporoparietal Junction in Autism Spectrum Disorder: A Multi-Site Study

Social function impairment is the core deficit of autism spectrum disorder (ASD). Although many studies have investigated ASD through a variety of neuroimaging tools, its brain mechanism of social function remains unclear due to its complex and heterogeneous symptoms. The present study aimed to use resting-state functional magnetic imaging data to explore effective connectivity between the right temporoparietal junction (RTPJ), one of the key brain regions associated with social impairment of individuals with ASD, and the whole brain to further deepen our understanding of the neuropathological mechanism of ASD. This study involved 1,454 participants from 23 sites from the Autism Brain Imaging Data Exchange (ABIDE) public dataset, which included 618 individuals with ASD and 836 with typical development (TD). First, a voxel-wise Granger causality analysis (GCA) was conducted with the RTPJ selected as the region of interest (ROI) to investigate the differences in effective connectivity between the ASD and TD groups in every site. Next, to obtain further accurate and representative results, an image-based meta-analysis was implemented to further analyze the GCA results of each site. Our results demonstrated abnormal causal connectivity between the RTPJ and the widely distributed brain regions and that the connectivity has been associated with social impairment in individuals with ASD. The current study could help to further elucidate the pathological mechanisms of ASD and provides a new perspective for future research.

Different functional connectivity optimal frequency in autism compared with healthy controls and the relationship with social communication deficits: Evidence from gene expression and behavior symptom analyses

Studies have reported that different brain regions/connections possess distinct frequency properties, which are related to brain function. Previous studies have proposed altered brain activity frequency and frequency-specific functional connectivity (FC) patterns in autism spectrum disorder (ASD), implying the varied dominant frequency of FC in ASD. However, the difference of the dominant frequency of FC between ASD and healthy controls (HCs) remains unclear. In the present study, the dominant frequency of FC was measured by FC optimal frequency, which was defined as the intermediate of the frequency bin at which the FC strength could reach the maximum. A multivariate pattern analysis was conducted to determine whether the FC optimal frequency in ASD differs from that in HCs. Partial least squares regression (PLSR) and enrichment analyses were conducted to determine the relationship between the FC optimal frequency difference of ASD/HCs and cortical gene expression. PLSR analyses were also performed to explore the relationship between FC optimal frequency and the clinical symptoms of ASD. Results showed a significant difference of FC optimal frequency between ASD and HCs. Some genes whose cortical expression patterns are related to the FC optimal frequency difference of ASD/HCs were enriched for social communication problems. Meanwhile, the FC optimal frequency in ASD was significantly related to social communication symptoms. These results may help us understand the neuro-mechanism of the social communication deficits in ASD.

Sparse Hierarchical Representation Learning on Functional Brain Networks for Prediction of Autism Severity Levels

Machine learning algorithms have been widely applied in diagnostic tools for autism spectrum disorder (ASD), revealing an altered brain connectivity. However, little is known about whether an magnetic resonance imaging (MRI)-based brain network is related to the severity of ASD symptoms in a large-scale cohort. We propose a graph convolution neural network-based framework that can generate sparse hierarchical graph representations for functional brain connectivity. Instead of assigning initial features for each node, we utilized a feature extractor to derive node features and the extracted representations can be fed to a hierarchical graph self-attention framework to effectively represent the entire graph. By incorporating connectivity embeddings in the feature extractor, we propose adjacency embedding networks to characterize the heterogeneous representations of the brain connectivity. Our proposed model variants outperform the benchmarking model with different configurations of adjacency embedding networks and types of functional connectivity matrices. Using this approach with the best configuration (SHEN atlas for node definition, Tikhonov correlation for connectivity estimation, and identity-adjacency embedding), we were able to predict individual ASD severity levels with a meaningful accuracy: the mean absolute error (MAE) and correlation between predicted and observed ASD severity scores resulted in 0.96, and r = 0.61 (P < 0.0001), respectively. To obtain a better understanding on how to generate better representations, we investigate the relationships between the extracted feature embeddings and the graph theory-based nodal measurements using canonical correlation analysis. Finally, we visualized the model to identify the most contributive functional connections for predicting ASD severity scores.

Identification of Young High-Functioning Autism Individuals Based on Functional Connectome Using Graph Isomorphism Network: A Pilot Study

Accumulated studies have determined the changes in functional connectivity in autism spectrum disorder (ASD) and spurred the application of machine learning for classifying ASD. Graph Neural Network provides a new method for network analysis in brain disorders to identify the underlying network features associated with functional deficits. Here, we proposed an improved model of Graph Isomorphism Network (GIN) that implements the Weisfeiler-Lehman (WL) graph isomorphism test to learn the graph features while taking into account the importance of each node in the classification to improve the interpretability of the algorithm. We applied the proposed method on multisite datasets of resting-state functional connectome from Autism Brain Imaging Data Exchange (ABIDE) after stringent quality control. The proposed method outperformed other commonly used classification methods on five different evaluation metrics. We also identified salient ROIs in visual and frontoparietal control networks, which could provide potential neuroimaging biomarkers for ASD identification.

Hyperbolic disc embedding of functional human brain connectomes using resting-state fMRI

The brain presents a real complex network of modular, small-world, and hierarchical nature, which are features of non-Euclidean geometry. Using resting-state functional magnetic resonance imaging, we constructed a scale-free binary graph for each subject, using internodal time series correlation of regions of interest as a proximity measure. The resulting network could be embedded onto manifolds of various curvatures and dimensions. While maintaining the fidelity of embedding (low distortion, high mean average precision), functional brain networks were found to be best represented in the hyperbolic disc. Using the 𝕊¹/ℍ² model, we reduced the dimension of the network into two-dimensional hyperbolic space and were able to efficiently visualize the internodal connections of the brain, preserving proximity as distances and angles on the hyperbolic discs. Each individual disc revealed relevance with its anatomic counterpart and absence of center-spaced node. Using the hyperbolic distance on the 𝕊¹/ℍ² model, we could detect the anomaly of network in autism spectrum disorder subjects. This procedure of embedding grants us a reliable new framework for studying functional brain networks and the possibility of detecting anomalies of the network in the hyperbolic disc on an individual scale.

Video-evoked fMRI BOLD responses are highly consistent across different data acquisition sites

Naturalistic imaging paradigms, in which participants view complex videos in the scanner, are increasingly used in human cognitive neuroscience. Videos evoke temporally synchronized brain responses that are similar across subjects as well as within subjects, but the reproducibility of these brain responses across different data acquisition sites has not yet been quantified. Here, we characterize the consistency of brain responses across independent samples of participants viewing the same videos in functional magnetic resonance imaging (fMRI) scanners at different sites (Indiana University and Caltech). We compared brain responses collected at these different sites for two carefully matched datasets with identical scanner models, acquisition, and preprocessing details, along with a third unmatched dataset in which these details varied. Our overall conclusion is that for matched and unmatched datasets alike, video-evoked brain responses have high consistency across these different sites, both when compared across groups and across pairs of individuals. As one might expect, differences between sites were larger for unmatched datasets than matched datasets. Residual differences between datasets could in part reflect participant-level variability rather than scanner- or data- related effects. Altogether our results indicate promise for the development and, critically, generalization of video fMRI studies of individual differences in healthy and clinical populations alike.

A Chinese multi-modal neuroimaging data release for increasing diversity of human brain mapping

The big-data use is becoming a standard practice in the neuroimaging field through data-sharing initiatives. It is important for the community to realize that such open science effort must protect personal, especially facial information when raw neuroimaging data are shared. An ideal tool for the face anonymization should not disturb subsequent brain tissue extraction and further morphological measurements. Using the high-resolution head images from magnetic resonance imaging (MRI) of 215 healthy Chinese, we discovered and validated a template effect on the face anonymization. Improved facial anonymization was achieved when the Chinese head templates but not the Western templates were applied to obscure the faces of Chinese brain images. This finding has critical implications for international brain imaging data-sharing. To facilitate the further investigation of potential culture-related impacts on and increase diversity of data-sharing for the human brain mapping, we released the 215 Chinese multi-modal MRI data into a database for imaging Chinese young brains, namely'I See your Brains (ISYB)', to the public via the Science Data Bank ( https://doi.org/10.11922/sciencedb.00740 ).

Association between resting-state functional brain connectivity and gene expression is altered in autism spectrum disorder

Gene expression covaries with brain activity as measured by resting state functional magnetic resonance imaging (MRI). However, it is unclear how genomic differences driven by disease state can affect this relationship. Here, we integrate from the ABIDE I and II imaging cohorts with datasets of gene expression in brains of neurotypical individuals and individuals with autism spectrum disorder (ASD) with regionally matched brain activity measurements from fMRI datasets. We identify genes linked with brain activity whose association is disrupted in ASD. We identified a subset of genes that showed a differential developmental trajectory in individuals with ASD compared with controls. These genes are enriched in voltage-gated ion channels and inhibitory neurons, pointing to excitation-inhibition imbalance in ASD. We further assessed differences at the regional level showing that the primary visual cortex is the most affected region in ASD. Our results link disrupted brain expression patterns of individuals with ASD to brain activity and show developmental, cell type, and regional enrichment of activity linked genes.

Multi-site harmonization of MRI data uncovers machine-learning discrimination capability in barely separable populations: An example from the ABIDE dataset

Machine Learning (ML) techniques have been widely used in Neuroimaging studies of Autism Spectrum Disorders (ASD) both to identify possible brain alterations related to this condition and to evaluate the predictive power of brain imaging modalities. The collection and public sharing of large imaging samples has favored an even greater diffusion of the use of ML-based analyses. However, multi-center data collections may suffer the batch effect, which, especially in case of Magnetic Resonance Imaging (MRI) studies, should be curated to avoid confounding effects for ML classifiers and masking biases. This is particularly important in the study of barely separable populations according to MRI data, such as subjects with ASD compared to controls with typical development (TD). Here, we show how the implementation of a harmo- nization protocol on brain structural features unlocks the case-control ML separation capability in the analysis of a multi-center MRI dataset. This effect is demonstrated on the ABIDE data collection, involving subjects encompassing a wide age range. After data harmonization, the overall ASD vs. TD discrimination capability by a Random Forest (RF) classifier improves from a very low performance (AUC = 0.58 ± 0.04) to a still low, but reasonably significant AUC = 0.67 ± 0.03. The performances of the RF classifier have been evaluated also in the age-specific subgroups of children, adolescents and adults, obtaining AUC = 0.62 ± 0.02, AUC = 0.65 ± 0.03 and AUC = 0.69 ± 0.06, respectively. Specific and consistent patterns of anatomical differences related to the ASD condition have been identified for the three different age subgroups.

A Brief Survey on No-Reference Image Quality Assessment Methods for Magnetic Resonance Images

No-reference image quality assessment (NR-IQA) methods automatically and objectively predict the perceptual quality of images without access to a reference image. Therefore, due to the lack of pristine images in most medical image acquisition systems, they play a major role in supporting the examination of resulting images and may affect subsequent treatment. Their usage is particularly important in magnetic resonance imaging (MRI) characterized by long acquisition times and a variety of factors that influence the quality of images. In this work, a survey covering recently introduced NR-IQA methods for the assessment of MR images is presented. First, typical distortions are reviewed and then popular NR methods are characterized, taking into account the way in which they describe MR images and create quality models for prediction. The survey also includes protocols used to evaluate the methods and popular benchmark databases. Finally, emerging challenges are outlined along with an indication of the trends towards creating accurate image prediction models.

Contrastive machine learning reveals the structure of neuroanatomical variation within autism

Autism spectrum disorder (ASD) is highly heterogeneous. Identifying systematic individual differences in neuroanatomy could inform diagnosis and personalized interventions. The challenge is that these differences are entangled with variation because of other causes: individual differences unrelated to ASD and measurement artifacts. We used contrastive deep learning to disentangle ASD-specific neuroanatomical variation from variation shared with typical control participants. ASD-specific variation correlated with individual differences in symptoms. The structure of this ASD-specific variation also addresses a long-standing debate about the nature of ASD: At least in terms of neuroanatomy, individuals do not cluster into distinct subtypes; instead, they are organized along continuous dimensions that affect distinct sets of regions.

Mapping the Heterogeneous Brain Structural Phenotype of Autism Spectrum Disorder Using the Normative Model

BACKGROUND

Autism spectrum disorder (ASD) is a complex neurodevelopmental disorder characterized by substantial clinical and biological heterogeneity. Quantitative and individualized metrics for delineating the heterogeneity of brain structure in ASD are still lacking. Likewise, the extent to which brain structural metrics of ASD deviate from typical development (TD) and whether deviations can be used for parsing brain structural phenotypes of ASD is unclear.

METHODS

T1-weighted magnetic resonance imaging data from the Autism Brain Imaging Data Exchange (ABIDE) II (n_TD = 564) were used to generate a normative model to map brain structure deviations of ABIDE I subjects (n_TD = 560, n_ASD = 496). Voxel-based morphometry was used to compute gray matter volume. Non-negative matrix factorization was employed to decompose the gray matter matrix into 6 factors and weights. These weights were used for normative modeling to estimate the factor deviations. Then, clustering analysis was used to identify ASD subtypes.

RESULTS

Compared with TD, ASD showed increased weights and deviations in 5 factors. Three subtypes with distinct neuroanatomical deviation patterns were identified. ASD subtype 1 and subtype 3 showed positive deviations, whereas ASD subtype 2 showed negative deviations. Distinct clinical manifestations in social communication deficits were identified among the three subtypes.

CONCLUSIONS

Our findings suggest that individuals with ASD have heterogeneous deviation patterns in brain structure. The results highlight the need to test for subtypes in neuroimaging studies of ASD. This study also presents a framework for understanding neuroanatomical heterogeneity in this increasingly prevalent neurodevelopmental disorder.

Alterations in Connectome Dynamics in Autism Spectrum Disorder: A Harmonized Mega- and Meta-analysis Study Using the Autism Brain Imaging Data Exchange Dataset

BACKGROUND

Neuroimaging studies have reported functional connectome aberrancies in autism spectrum disorder (ASD). However, the time-varying patterns of connectome topology in individuals with ASD and the connection between these patterns and gene expression profiles remain unknown.

METHODS

To investigate case-control differences in dynamic connectome topology, we conducted mega- and meta-analyses of resting-state functional magnetic resonance imaging data of 939 participants (440 patients with ASD and 499 healthy control subjects, all males) from 18 independent sites, selected from the Autism Brain Imaging Data Exchange (ABIDE) dataset. Functional data were preprocessed and analyzed using harmonized protocols, and brain module dynamics was assessed using a multilayer network model. We further leveraged postmortem brain-wide gene expression data to identify transcriptomic signatures associated with ASD-related alterations in brain dynamics.

RESULTS

Compared with healthy control participants, individuals with ASD exhibited a higher global mean and lower standard deviation of whole-brain module dynamics, indicating an unstable and less regionally differentiated pattern. More specifically, individuals with ASD showed higher module switching, primarily in the medial prefrontal cortex, posterior cingulate gyrus, and angular gyrus, and lower switching in the visual regions. These alterations in brain dynamics were predictive of social impairments in individuals with ASD and were linked with expression profiles of genes primarily involved in the regulation of neurotransmitter transport and secretion as well as with previously identified autism-related genes.

CONCLUSIONS

This study is the first to identify consistent alterations in brain network dynamics in ASD and the transcriptomic signatures related to those alterations, furthering insights into the biological basis behind this disorder.

Identification of Pathogenetic Brain Regions via Neuroimaging Data for Diagnosis of Autism Spectrum Disorders

Autism spectrum disorder (ASD) is a kind of neurodevelopmental disorder that often occurs in children and has a hidden onset. Patients usually have lagged development of communication ability and social behavior and thus suffer an unhealthy physical and mental state. Evidence has indicated that diseases related to ASD have commonalities in brain imaging characteristics. This study aims to study the pathogenesis of ASD based on brain imaging data to locate the ASD-related brain regions. Specifically, we collected the functional magnetic resonance image data of 479 patients with ASD and 478 normal subjects matched in age and gender and used a machine-learning framework named random support vector machine cluster to extract distinctive brain regions from the preprocessed data. According to the experimental results, compared with other existing approaches, the method used in this study can more accurately distinguish patients from normal individuals based on brain imaging data. At the same time, this study found that the development of ASD was highly correlated with certain brain regions, e.g., lingual gyrus, superior frontal gyrus, medial gyrus, insular lobe, and olfactory cortex. This study explores the effectiveness of a novel machine-learning approach in the study of ASD brain imaging and provides a reference brain area for the medical research and clinical treatment of ASD.

FastSurferVINN: Building resolution-independence into deep learning segmentation methods-A solution for HighRes brain MRI

Leading neuroimaging studies have pushed 3T MRI acquisition resolutions below 1.0 mm for improved structure definition and morphometry. Yet, only few, time-intensive automated image analysis pipelines have been validated for high-resolution (HiRes) settings. Efficient deep learning approaches, on the other hand, rarely support more than one fixed resolution (usually 1.0 mm). Furthermore, the lack of a standard submillimeter resolution as well as limited availability of diverse HiRes data with sufficient coverage of scanner, age, diseases, or genetic variance poses additional, unsolved challenges for training HiRes networks. Incorporating resolution-independence into deep learning-based segmentation, i.e., the ability to segment images at their native resolution across a range of different voxel sizes, promises to overcome these challenges, yet no such approach currently exists. We now fill this gap by introducing a Voxel-size Independent Neural Network (VINN) for resolution-independent segmentation tasks and present FastSurferVINN, which (i) establishes and implements resolution-independence for deep learning as the first method simultaneously supporting 0.7-1.0 mm whole brain segmentation, (ii) significantly outperforms state-of-the-art methods across resolutions, and (iii) mitigates the data imbalance problem present in HiRes datasets. Overall, internal resolution-independence mutually benefits both HiRes and 1.0 mm MRI segmentation. With our rigorously validated FastSurferVINN we distribute a rapid tool for morphometric neuroimage analysis. The VINN architecture, furthermore, represents an efficient resolution-independent segmentation method for wider application.

rs-fMRI and machine learning for ASD diagnosis: a systematic review and meta-analysis

Autism Spectrum Disorder (ASD) diagnosis is still based on behavioral criteria through a lengthy and time-consuming process. Much effort is being made to identify brain imaging biomarkers and develop tools that could facilitate its diagnosis. In particular, using Machine Learning classifiers based on resting-state fMRI (rs-fMRI) data is promising, but there is an ongoing need for further research on their accuracy and reliability. Therefore, we conducted a systematic review and meta-analysis to summarize the available evidence in the literature so far. A bivariate random-effects meta-analytic model was implemented to investigate the sensitivity and specificity across the 55 studies that offered sufficient information for quantitative analysis. Our results indicated overall summary sensitivity and specificity estimates of 73.8% and 74.8%, respectively. SVM stood out as the most used classifier, presenting summary estimates above 76%. Studies with bigger samples tended to obtain worse accuracies, except in the subgroup analysis for ANN classifiers. The use of other brain imaging or phenotypic data to complement rs-fMRI information seems promising, achieving higher sensitivities when compared to rs-fMRI data alone (84.7% versus 72.8%). Finally, our analysis showed AUC values between acceptable and excellent. Still, given the many limitations indicated in our study, further well-designed studies are warranted to extend the potential use of those classification algorithms to clinical settings.

A Fast Transform for Brain Connectivity Difference Evaluation

Anatomical and dynamical connectivity are essential to healthy brain function. However, quantifying variations in connectivity across conditions or between patient populations and appraising their functional significance are highly non-trivial tasks. Here we show that link ranking differences induce specific geometries in a convenient auxiliary space that are often easily recognisable at mere eye inspection. Link ranking can also provide fast and reliable criteria for network reconstruction parameters for which no theoretical guideline has been proposed.

Population heterogeneity in clinical cohorts affects the predictive accuracy of brain imaging

Brain imaging research enjoys increasing adoption of supervised machine learning for single-participant disease classification. Yet, the success of these algorithms likely depends on population diversity, including demographic differences and other factors that may be outside of primary scientific interest. Here, we capitalize on propensity scores as a composite confound index to quantify diversity due to major sources of population variation. We delineate the impact of population heterogeneity on the predictive accuracy and pattern stability in 2 separate clinical cohorts: the Autism Brain Imaging Data Exchange (ABIDE, n = 297) and the Healthy Brain Network (HBN, n = 551). Across various analysis scenarios, our results uncover the extent to which cross-validated prediction performances are interlocked with diversity. The instability of extracted brain patterns attributable to diversity is located preferentially in regions part of the default mode network. Collectively, our findings highlight the limitations of prevailing deconfounding practices in mitigating the full consequences of population diversity.

Subcortical Brain Morphometry Differences between Adults with Autism Spectrum Disorder and Schizophrenia

Autism spectrum disorder (ASD) and schizophrenia (SZ) are neuropsychiatric disorders that overlap in symptoms associated with social-cognitive impairment. Subcortical structures play a significant role in cognitive and social-emotional behaviors and their abnormalities are associated with neuropsychiatric conditions. This exploratory study utilized ABIDE II/COBRE MRI and corresponding phenotypic datasets to compare subcortical volumes of adults with ASD (n = 29), SZ (n = 51) and age and gender matched neurotypicals (NT). We examined the association between subcortical volumes and select behavioral measures to determine whether core symptomatology of disorders could be explained by subcortical association patterns. We observed volume differences in ASD (viz., left pallidum, left thalamus, left accumbens, right amygdala) but not in SZ compared to their respective NT controls, reflecting morphometric changes specific to one of the disorder groups. However, left hippocampus and amygdala volumes were implicated in both disorders. A disorder-specific negative correlation (r = −0.39, p = 0.038) was found between left-amygdala and scores on the Social Responsiveness Scale (SRS) Social-Cognition in ASD, and a positive association (r = 0.29, p = 0.039) between full scale IQ (FIQ) and right caudate in SZ. Significant correlations between behavior measures and subcortical volumes were observed in NT groups (ASD-NT range; r = −0.53 to −0.52, p = 0.002 to 0.004, SZ-NT range; r = −0.41 to −0.32, p = 0.007 to 0.021) that were non-significant in the disorder groups. The overlap of subcortical volumes implicated in ASD and SZ may reflect common neurological mechanisms. Furthermore, the difference in correlation patterns between disorder and NT groups may suggest dysfunctional connectivity with cascading effects unique to each disorder and a potential role for IQ in mediating behavior and brain circuits.