Visualization and unsupervised predictive clustering of high-dimensional multimodal neuroimaging data

Individualized Assessment of Brain Aβ Deposition With fMRI Using Deep Learning

PET-based Alzheimer's disease (AD) assessment has many limitations in large-scale screening. Non-invasive techniques such as resting-state functional magnetic resonance imaging (rs-fMRI) have been proven valuable in early AD diagnosis. This study investigated feasibility of using rs-fMRI, especially functional connectivity (FC), for individualized assessment of brain amyloid-β deposition derived from PET. We designed a graph convolutional networks (GCNs) and random forest (RF) based integrated framework for using rs-fMRI-derived multi-level FC networks to predict amyloid-β PET patterns with the OASIS-3 (N = 258) and ADNI-2 (N = 291) datasets. Our method achieved satisfactory accuracy not only in Aβ-PET grade classification (for negative, intermediate, and positive grades, with accuracy in the three-class classification as 62.8% and 64.3% on two datasets, respectively), but also in prediction of whole-brain region-level Aβ-PET standard uptake value ratios (SUVRs) (with the mean square errors as 0.039 and 0.074 for two datasets, respectively). Model interpretability examination also revealed the contributive role of the limbic network. This study demonstrated high feasibility and reproducibility of using low-cost, more accessible magnetic resonance imaging (MRI) to approximate PET-based diagnosis.

Biotypes of major depressive disorder identified by a multiview clustering framework

BACKGROUND

The advances in resting-state functional magnetic resonance imaging techniques motivate parsing heterogeneity in major depressive disorder (MDD) through neurophysiological subtypes (i.e., biotypes). Based on graph theories, researchers have observed the functional organization of the human brain as a complex system with modular structures and have found wide-spread but variable MDD-related abnormality regarding the modules. The evidence implies the possibility of identifying biotypes using high-dimensional functional connectivity (FC) data in ways that suit the potentially multifaceted biotypes taxonomy.

METHODS

We proposed a multiview biotype discovery framework that involves theory-driven feature subspace partition (i.e., "view") and independent subspace clustering. Six views were defined using intra- and intermodule FC regarding three MDD focal modules (i.e., the sensory-motor system, default mode network, and subcortical network). For robust biotypes, the framework was applied to a large multisite sample (805 MDD participants and 738 healthy controls).

RESULTS

Two biotypes were stably obtained in each view, respectively characterized by significantly increased and decreased FC compared to healthy controls. These view-specific biotypes promoted the diagnosis of MDD and showed different symptom profiles. By integrating the view-specific biotypes into biotype profiles, a broad spectrum in the neural heterogeneity of MDD and its separation from symptom-based subtypes was further revealed.

LIMITATIONS

The power of clinical effects is limited and the cross-sectional nature cannot predict the treatment effects of the biotypes.

CONCLUSIONS

Our findings not only contribute to the understanding of heterogeneity in MDD, but also provide a novel subtyping framework that could transcend current diagnostic boundaries and data modality.

Predicting Brain Amyloid-β PET Grades with Graph Convolutional Networks Based on Functional MRI and Multi-Level Functional Connectivity

BACKGROUND

The detection of amyloid-β (Aβ) deposition in the brain provides crucial evidence in the clinical diagnosis of Alzheimer's disease (AD). However, the current positron emission tomography (PET)-based brain Aβ examination suffers from the problems of coarse visual inspection (in many cases, with 2-class stratification) and high scanning cost.

OBJECTIVE

1) To characterize the non-binary Aβ deposition levels in the AD continuum based on clustering of PET data, and 2) to explore the feasibility of predicting individual Aβ deposition grades with non-invasive functional magnetic resonance imaging (fMRI).

METHODS

1) Individual whole-brain Aβ-PET images from the OASIS-3 dataset (N = 258) were grouped into three clusters (grades) with t-SNE and k-means. The demographical data as well as global and regional standard uptake value ratios (SUVRs) were compared among the three clusters with Chi-square tests or ANOVA tests. 2) From resting-state fMRI, both conventional functional connectivity (FC) and high-order FC networks were constructed and the topological architectures of the two networks were jointly learned with graph convolutional networks (GCNs) to predict the Aβ-PET grades for each individual.

RESULTS

We found three clearly separated clusters, indicating three Aβ-PET grades. There were significant differences in gender, age, cognitive ability, APOE type, as well as global and regional SUVRs among the three grades we found. The prediction of Aβ-PET grades with GCNs on FC for the 258 participants in the AD continuum reached a satisfactory averaged accuracy (78.8%) in the two-class classification tasks.

CONCLUSION

The results demonstrated the feasibility of using deep learning on a non-invasive brain functional imaging technique to approximate PET-based Aβ deposition grading.

Genome-wide analysis of therapeutic response uncovers molecular pathways governing tamoxifen resistance in ER+ breast cancer

Background

Prioritization of breast cancer patients based on the risk of resistance to tamoxifen plays a significant role in personalized therapeutic planning and improving disease course and outcomes.

Methods

In this work, we demonstrate that a genome-wide pathway-centric computational framework elucidates molecular pathways as markers of tamoxifen resistance in ER+ breast cancer patients. In particular, we associated activity levels of molecular pathways with a wide spectrum of response to tamoxifen, which defined markers of tamoxifen resistance in patients with ER+ breast cancer.

Findings

We identified five biological pathways as markers of tamoxifen failure and demonstrated their ability to predict the risk of tamoxifen resistance in two independent patient cohorts (Test cohort1: log-rank p-value = 0.02, adjusted HR = 3.11; Test cohort2: log-rank p-value = 0.01, adjusted HR = 4.24). We have shown that these pathways are not markers of aggressiveness and outperform known markers of tamoxifen response. Furthermore, for adoption into clinic, we derived a list of pathway read-out genes and their associated scoring system, which assigns a risk of tamoxifen resistance for new incoming patients.

Interpretation

We propose that the identified pathways and their read-out genes can be utilized to prioritize patients who would benefit from tamoxifen treatment and patients at risk of tamoxifen resistance that should be offered alternative regimens.

Funding

This work was supported by the Rutgers SHP Dean's research grant, Rutgers start-up funds, Libyan Ministry of Higher Education and Scientific Research, and Katrina Kehlet Graduate Award from The NJ Chapter of the Healthcare Information Management Systems Society.

Hypothalamic estrogen receptor alpha establishes a sexually dimorphic regulatory node of energy expenditure

Estrogen receptor a (ERa) signaling in the ventromedial hypothalamus (VMH) contributes to energy homeostasis by modulating physical activity and thermogenesis. However, the precise neuronal populations involved remain undefined. Here, we describe six neuronal populations in the mouse VMH by using single-cell RNA transcriptomics and in situ hybridization. ERa is enriched in populations showing sex biased expression of reprimo (Rprm), tachykinin 1 (Tac1), and prodynorphin (Pdyn). Female biased expression of Tac1 and Rprm is patterned by ERa-dependent repression during male development, whereas male biased expression of Pdyn is maintained by circulating testicular hormone in adulthood. Chemogenetic activation of ERa positive VMH neurons stimulates heat generation and movement in both sexes. However, silencing Rprm gene function increases core temperature selectively in females and ectopic Rprm expression in males is associated with reduced core temperature. Together these findings reveal a role for Rprm in temperature regulation and ERa in the masculinization of neuron populations that underlie energy expenditure.

Requirements of Data Visualisation Tools to Analyse Big Data: A Structured Literature Review

The continual growth of big data necessitates efficient ways of analysing these large datasets. Data visualisation and visual analytics has been identified as a key tool in big data analysis because they draw on the human visual and cognitive capabilities to analyse data quickly, intuitively and interactively. However, current visualisation tools and visual analytical systems fall short of providing a seamless user experience and several improvements could be made to current commercially available visualisation tools. By conducting a systematic literature review, requirements of visualisation tools were identified and categorised into six groups: dimensionality reduction, data reduction, scalability and readability, interactivity, fast retrieval of results, and user assistance. The most common themes found in the literature were dimensionality reduction and interactive data exploration.

Heterogeneous beta-catenin activation is sufficient to cause hepatocellular carcinoma in zebrafish

Up to 41% of hepatocellular carcinomas (HCCs) result from activating mutations in the CTNNB1 gene encoding β-catenin. HCC-associated CTNNB1 mutations stabilize the β-catenin protein, leading to nuclear and/or cytoplasmic localization of β-catenin and downstream activation of Wnt target genes. In patient HCC samples, β-catenin nuclear and cytoplasmic localization are typically patchy, even among HCC with highly active CTNNB1 mutations. The functional and clinical relevance of this heterogeneity in β-catenin activation are not well understood. To define mechanisms of β-catenin-driven HCC initiation, we generated a Cre-lox system that enabled switching on activated β-catenin in (1) a small number of hepatocytes in early development; or (2) the majority of hepatocytes in later development or adulthood. We discovered that switching on activated β-catenin in a subset of larval hepatocytes was sufficient to drive HCC initiation. To determine the role of Wnt/β-catenin signaling heterogeneity later in hepatocarcinogenesis, we performed RNA-seq analysis of zebrafish β-catenin-driven HCC. At the single-cell level, 2.9% to 15.2% of hepatocytes from zebrafish β-catenin-driven HCC expressed two or more of the Wnt target genes axin2, mtor, glula, myca and wif1, indicating focal activation of Wnt signaling in established tumors. Thus, heterogeneous β-catenin activation drives HCC initiation and persists throughout hepatocarcinogenesis.

Tracking the Brain State Transition Process of Dynamic Function Connectivity Based on Resting State fMRI

BOLD-fMRI technology provides a good foundation for the research of human brain dynamic functional connectivity and brain state analysis. However, due to the complexity of brain function connectivity and the high dimensionality expression of brain dynamic attributions, more research studies are focusing on tracking the time-varying characteristics through the transition between different brain states. The transition process is considered to occur instantaneously at some special time point in the above research studies, whereas our work found the brain state transition may be completed in a time section gradually rather than instantaneously. In this paper, a brain state conversion rate model is constructed to observe the procedure of brain state transition trend at each time point, and the state change can be observed by the values of conversion rate. According to the results, the transition of status always lasts for a few time points, and a brain state network model with both steady state and transition state is presented. Network topological overlap coefficient is built to analyze the features of time-varying networks. With this method, some common regular patterns of time-varying characteristics can be observed strongly in healthy children but not in the autism children. This distinct can help us to distinguish children with autism from healthy children.

Automated cortical thickness and skewness feature selection in bipolar disorder using a semi-supervised learning method

BACKGROUND

Bipolar disorder (BD) broadly affects brain structure, in particular areas involved in emotion processing and cognition. In the last years, the psychiatric field's interest in machine learning approaches has been steadily growing, thanks to the potentiality of automatically discriminating patients from healthy controls.

METHODS

In this work, we employed cortical thickness of 58 regions of interest obtained from magnetic resonance imaging scans of 41 BD patients and 34 healthy controls, to automatically identify the regions which are mostly involved with the disease. We used a semi-supervised method, addressing the criticisms on supervised methods, related to the fact that the diagnosis is not unaffected by uncertainty.

RESULTS

Our results confirm findings in previous studies, with a classification accuracy of about 75% when mean thickness and skewness of up to five regions are considered. We obtained that the parietal lobe and some areas in the temporal sulcus were the regions which were the most involved with BD.

LIMITATIONS

The major limitation of our work is the limited size or our dataset, but in line with other recent machine learning works in the field. Moreover, we considered chronic patients, whose brain characteristics may thus be affected.

CONCLUSIONS

The automatic selection of the brain regions most involved in BD may be of great importance when dealing with the pathogenesis of the disorder. Our method selected regions which are known to be involved with BD, indicating that damage to the identified areas can be considered as a marker of disease.

Multimodal Neuroimaging: Basic Concepts and Classification of Neuropsychiatric Diseases

Neuroimaging techniques are widely used in neuroscience to visualize neural activity, to improve our understanding of brain mechanisms, and to identify biomarkers-especially for psychiatric diseases; however, each neuroimaging technique has several limitations. These limitations led to the development of multimodal neuroimaging (MN), which combines data obtained from multiple neuroimaging techniques, such as electroencephalography, functional magnetic resonance imaging, and yields more detailed information about brain dynamics. There are several types of MN, including visual inspection, data integration, and data fusion. This literature review aimed to provide a brief summary and basic information about MN techniques (data fusion approaches in particular) and classification approaches. Data fusion approaches are generally categorized as asymmetric and symmetric. The present review focused exclusively on studies based on symmetric data fusion methods (data-driven methods), such as independent component analysis and principal component analysis. Machine learning techniques have recently been introduced for use in identifying diseases and biomarkers of disease. The machine learning technique most widely used by neuroscientists is classification-especially support vector machine classification. Several studies differentiated patients with psychiatric diseases and healthy controls with using combined datasets. The common conclusion among these studies is that the prediction of diseases increases when combining data via MN techniques; however, there remain a few challenges associated with MN, such as sample size. Perhaps in the future N-way fusion can be used to combine multiple neuroimaging techniques or nonimaging predictors (eg, cognitive ability) to overcome the limitations of MN.

Deep Learning in Neuroradiology

Deep learning is a form of machine learning using a convolutional neural network architecture that shows tremendous promise for imaging applications. It is increasingly being adapted from its original demonstration in computer vision applications to medical imaging. Because of the high volume and wealth of multimodal imaging information acquired in typical studies, neuroradiology is poised to be an early adopter of deep learning. Compelling deep learning research applications have been demonstrated, and their use is likely to grow rapidly. This review article describes the reasons, outlines the basic methods used to train and test deep learning models, and presents a brief overview of current and potential clinical applications with an emphasis on how they are likely to change future neuroradiology practice. Facility with these methods among neuroimaging researchers and clinicians will be important to channel and harness the vast potential of this new method.

Brain Subtyping Enhances The Neuroanatomical Discrimination of Schizophrenia

Identifying distinctive subtypes of schizophrenia could ultimately enhance diagnostic and prognostic accuracy. We aimed to uncover neuroanatomical subtypes of chronic schizophrenia patients to test whether stratification can enhance computer-aided discrimination of patients from control subjects. Unsupervised, data-driven clustering of structural MRI (sMRI) data was used to identify 2 subtypes of schizophrenia patients drawn from a US-based open science repository (n = 71) and we quantified classification improvements compared to controls (n = 74) using supervised machine learning. We externally validated the unsupervised and supervised learning models in a heterogeneous German validation sample (n = 316), and characterized symptom, cognition, and longitudinal symptom change signatures. Stratification improved classification accuracies from 68.5% to 73% (subgroup 1) and 78.8% (subgroup 2), respectively. Increased accuracy was also found when models were externally validated, and an average gain of 9% was found in supplementary analyses. The first subgroup was associated with cortical and subcortical volume reductions coupled with substantially longer illness duration, whereas the second subgroup was mainly characterized by cortical reductions, reduced illness duration, and comparatively less negative symptoms. Individuals within each subgroup could be identified using just 10 clinical questions at an accuracy of 81.2%, and differential cognitive and symptom course signatures were suggested in multivariate analyses. Our findings suggest that sMRI-based subtyping enhances the neuroanatomical discrimination of schizophrenia by identifying generalizable brain patterns that align with a clinical staging model of the disorder. These findings could be used to improve illness stratification for biomarker-based computer-aided diagnoses.

Moving Beyond ERP Components: A Selective Review of Approaches to Integrate EEG and Behavior

Relationships between neuroimaging measures and behavior provide important clues about brain function and cognition in healthy and clinical populations. While electroencephalography (EEG) provides a portable, low cost measure of brain dynamics, it has been somewhat underrepresented in the emerging field of model-based inference. We seek to address this gap in this article by highlighting the utility of linking EEG and behavior, with an emphasis on approaches for EEG analysis that move beyond focusing on peaks or “components” derived from averaging EEG responses across trials and subjects (generating the event-related potential, ERP). First, we review methods for deriving features from EEG in order to enhance the signal within single-trials. These methods include filtering based on user-defined features (i.e., frequency decomposition, time-frequency decomposition), filtering based on data-driven properties (i.e., blind source separation, BSS), and generating more abstract representations of data (e.g., using deep learning). We then review cognitive models which extract latent variables from experimental tasks, including the drift diffusion model (DDM) and reinforcement learning (RL) approaches. Next, we discuss ways to access associations among these measures, including statistical models, data-driven joint models and cognitive joint modeling using hierarchical Bayesian models (HBMs). We think that these methodological tools are likely to contribute to theoretical advancements, and will help inform our understandings of brain dynamics that contribute to moment-to-moment cognitive function.

Machine-Learning Algorithms to Automate Morphological and Functional Assessments in 2D Echocardiography

BACKGROUND

Machine-learning models may aid cardiac phenotypic recognition by using features of cardiac tissue deformation.

OBJECTIVES

This study investigated the diagnostic value of a machine-learning framework that incorporates speckle-tracking echocardiographic data for automated discrimination of hypertrophic cardiomyopathy (HCM) from physiological hypertrophy seen in athletes (ATH).

METHODS

Expert-annotated speckle-tracking echocardiographic datasets obtained from 77 ATH and 62 HCM patients were used for developing an automated system. An ensemble machine-learning model with 3 different machine-learning algorithms (support vector machines, random forests, and artificial neural networks) was developed and a majority voting method was used for conclusive predictions with further K-fold cross-validation.

RESULTS

Feature selection using an information gain (IG) algorithm revealed that volume was the best predictor for differentiating between HCM ands. ATH (IG = 0.24) followed by mid-left ventricular segmental (IG = 0.134) and average longitudinal strain (IG = 0.131). The ensemble machine-learning model showed increased sensitivity and specificity compared with early-to-late diastolic transmitral velocity ratio (p < 0.01), average early diastolic tissue velocity (e') (p < 0.01), and strain (p = 0.04). Because ATH were younger, adjusted analysis was undertaken in younger HCM patients and compared with ATH with left ventricular wall thickness >13 mm. In this subgroup analysis, the automated model continued to show equal sensitivity, but increased specificity relative to early-to-late diastolic transmitral velocity ratio, e', and strain.

CONCLUSIONS

Our results suggested that machine-learning algorithms can assist in the discrimination of physiological versus pathological patterns of hypertrophic remodeling. This effort represents a step toward the development of a real-time, machine-learning-based system for automated interpretation of echocardiographic images, which may help novice readers with limited experience.

Identification and individualized prediction of clinical phenotypes in bipolar disorders using neurocognitive data, neuroimaging scans and machine learning

Diagnosis, clinical management and research of psychiatric disorders remain subjective - largely guided by historically developed categories which may not effectively capture underlying pathophysiological mechanisms of dysfunction. Here, we report a novel approach of identifying and validating distinct and biologically meaningful clinical phenotypes of bipolar disorders using both unsupervised and supervised machine learning techniques. First, neurocognitive data were analyzed using an unsupervised machine learning approach and two distinct clinical phenotypes identified namely; phenotype I and phenotype II. Second, diffusion weighted imaging scans were pre-processed using the tract-based spatial statistics (TBSS) method and 'skeletonized' white matter fractional anisotropy (FA) and mean diffusivity (MD) maps extracted. The 'skeletonized' white matter FA and MD maps were entered into the Elastic Net machine learning algorithm to distinguish individual subjects' phenotypic labels (e.g. phenotype I vs. phenotype II). This calculation was performed to ascertain whether the identified clinical phenotypes were biologically distinct. Original neurocognitive measurements distinguished individual subjects' phenotypic labels with 94% accuracy (sensitivity=92%, specificity=97%). TBSS derived FA and MD measurements predicted individual subjects' phenotypic labels with 76% and 65% accuracy respectively. In addition, individual subjects belonging to phenotypes I and II were distinguished from healthy controls with 57% and 92% accuracy respectively. Neurocognitive task variables identified as most relevant in distinguishing phenotypic labels included; Affective Go/No-Go (AGN), Cambridge Gambling Task (CGT) coupled with inferior fronto-occipital fasciculus and callosal white matter pathways. These results suggest that there may exist two biologically distinct clinical phenotypes in bipolar disorders which can be identified from healthy controls with high accuracy and at an individual subject level. We suggest a strong clinical utility of the proposed approach in defining and validating biologically meaningful and less heterogeneous clinical sub-phenotypes of major psychiatric disorders.

Beyond Lumping and Splitting: A Review of Computational Approaches for Stratifying Psychiatric Disorders

Heterogeneity is a key feature of all psychiatric disorders that manifests on many levels, including symptoms, disease course, and biological underpinnings. These form a substantial barrier to understanding disease mechanisms and developing effective, personalized treatments. In response, many studies have aimed to stratify psychiatric disorders, aiming to find more consistent subgroups on the basis of many types of data. Such approaches have received renewed interest after recent research initiatives, such as the National Institute of Mental Health Research Domain Criteria and the European Roadmap for Mental Health Research, both of which emphasize finding stratifications that are based on biological systems and that cut across current classifications. We first introduce the basic concepts for stratifying psychiatric disorders and then provide a methodologically oriented and critical review of the existing literature. This shows that the predominant clustering approach that aims to subdivide clinical populations into more coherent subgroups has made a useful contribution but is heavily dependent on the type of data used; it has produced many different ways to subgroup the disorders we review, but for most disorders it has not converged on a consistent set of subgroups. We highlight problems with current approaches that are not widely recognized and discuss the importance of validation to ensure that the derived subgroups index clinically relevant variation. Finally, we review emerging techniques-such as those that estimate normative models for mappings between biology and behavior-that provide new ways to parse the heterogeneity underlying psychiatric disorders and evaluate all methods to meeting the objectives of such as the National Institute of Mental Health Research Domain Criteria and Roadmap for Mental Health Research.

A Tool for Interactive Data Visualization: Application to Over 10,000 Brain Imaging and Phantom MRI Data Sets

In this paper we propose a web-based approach for quick visualization of big data from brain magnetic resonance imaging (MRI) scans using a combination of an automated image capture and processing system, nonlinear embedding, and interactive data visualization tools. We draw upon thousands of MRI scans captured via the COllaborative Imaging and Neuroinformatics Suite (COINS). We then interface the output of several analysis pipelines based on structural and functional data to a t-distributed stochastic neighbor embedding (t-SNE) algorithm which reduces the number of dimensions for each scan in the input data set to two dimensions while preserving the local structure of data sets. Finally, we interactively display the output of this approach via a web-page, based on data driven documents (D3) JavaScript library. Two distinct approaches were used to visualize the data. In the first approach, we computed multiple quality control (QC) values from pre-processed data, which were used as inputs to the t-SNE algorithm. This approach helps in assessing the quality of each data set relative to others. In the second case, computed variables of interest (e.g., brain volume or voxel values from segmented gray matter images) were used as inputs to the t-SNE algorithm. This approach helps in identifying interesting patterns in the data sets. We demonstrate these approaches using multiple examples from over 10,000 data sets including (1) quality control measures calculated from phantom data over time, (2) quality control data from human functional MRI data across various studies, scanners, sites, (3) volumetric and density measures from human structural MRI data across various studies, scanners and sites. Results from (1) and (2) show the potential of our approach to combine t-SNE data reduction with interactive color coding of variables of interest to quickly identify visually unique clusters of data (i.e., data sets with poor QC, clustering of data by site) quickly. Results from (3) demonstrate interesting patterns of gray matter and volume, and evaluate how they map onto variables including scanners, age, and gender. In sum, the proposed approach allows researchers to rapidly identify and extract meaningful information from big data sets. Such tools are becoming increasingly important as datasets grow larger.

A group ICA based framework for evaluating resting fMRI markers when disease categories are unclear: application to schizophrenia, bipolar, and schizoaffective disorders

Schizophrenia (SZ), bipolar disorder (BP) and schizoaffective disorder (SAD) share some common symptoms, and there is still a debate about whether SAD is an independent category. To the best of our knowledge, no study has been done to differentiate these three disorders or to investigate the distinction of SAD as an independent category using fMRI data. This study is aimed to explore biomarkers from resting-state fMRI networks for differentiating these disorders and investigate the relationship among these disorders based on fMRI networks with an emphasis on SAD. Firstly, a novel group ICA method, group information guided independent component analysis (GIG-ICA), was applied to extract subject-specific brain networks from fMRI data of 20 healthy controls (HC), 20 SZ patients, 20 BP patients, 20 patients suffering from SAD with manic episodes (SADM), and 13 patients suffering from SAD with depressive episodes exclusively (SADD). Then, five-level one-way analysis of covariance and multiclass support vector machine recursive feature elimination were employed to identify discriminative regions from the networks. Subsequently, the t-distributed stochastic neighbor embedding (t-SNE) projection and the hierarchical clustering were implemented to investigate the relationship among those groups. Finally, to evaluate the generalization ability, 16 new subjects were classified based on the found regions and the trained model using original 93 subjects. Results show that the discriminative regions mainly included frontal, parietal, precuneus, cingulate, supplementary motor, cerebellar, insula and supramarginal cortices, which performed well in distinguishing different groups. SADM and SADD were the most similar to each other, although SADD had greater similarity to SZ compared to other groups, which indicates that SAD may be an independent category. BP was closer to HC compared with other psychotic disorders. In summary, resting-state fMRI brain networks extracted via GIG-ICA provide a promising potential to differentiate SZ, BP, and SAD.