A Comprehensive Framework for Differentiating Autism Spectrum Disorder From Neurotypicals by Fusing Structural MRI and Resting State Functional MRI

BACKGROUND Over the recent years, machine learning algorithms have been more widely and increasingly applied in biomedical fields. In particular, its application has been drawing more attention in the field of psychiatry, for instance, as diagnostic tests/tools for autism spectrum disorder. However, given its complexity and potential clinical implications, there is ongoing need for further research on its accuracy. OBJECTIVE The current study aims to summarize the evidence for the accuracy of use of machine learning algorithms in diagnosing autism spectrum disorder (ASD) through systematic review and meta-analysis. METHODS MEDLINE, Embase, CINAHL Complete (with OpenDissertations), PsyINFO and IEEE Xplore Digital Library databases were searched on November 28th, 2018. Studies, which used a machine learning algorithm partially or fully in classifying ASD from controls and provided accuracy measures, were included in our analysis. Bivariate random effects model was applied to the pooled data in meta-analysis. Subgroup analysis was used to investigate and resolve the source of heterogeneity between studies. True-positive, false-positive, false negative and true-negative values from individual studies were used to calculate the pooled sensitivity and specificity values, draw SROC curves, and obtain area under the curve (AUC) and partial AUC. RESULTS A total of 43 studies were included for the final analysis, of which meta-analysis was performed on 40 studies (53 samples with 12,128 participants). A structural MRI subgroup meta-analysis (12 samples with 1,776 participants) showed the sensitivity at 0.83 (95% CI-0.76 to 0.89), specificity at 0.84 (95% CI -0.74 to 0.91), and AUC/pAUC at 0.90/0.83. An fMRI/deep neural network (DNN) subgroup meta-analysis (five samples with 1,345 participants) showed the sensitivity at 0.69 (95% CI- 0.62 to 0.75), the specificity at 0.66 (95% CI -0.61 to 0.70), and AUC/pAUC at 0.71/0.67. CONCLUSIONS Machine learning algorithms that used structural MRI features in diagnosis of ASD were shown to have accuracy that is similar to currently used diagnostic tools.

Download Full-text

Brain-based sex differences in autism spectrum disorder across the lifespan: A systematic review of structural MRI, fMRI, and DTI findings

NeuroImage Clinical ◽

10.1016/j.nicl.2021.102719 ◽

2021 ◽

pp. 102719

Author(s):

Melissa J.M. Walsh ◽

Gregory L. Wallace ◽

Stephen M. Gallegos ◽

B. Blair Braden

Keyword(s):

Systematic Review ◽

Autism Spectrum Disorder ◽

Sex Differences ◽

Structural Mri ◽

Autism Spectrum ◽

Spectrum Disorder

Download Full-text

A Deep Learning Approach to Predict Autism Spectrum Disorder Using Multisite Resting-State fMRI

Applied Sciences ◽

10.3390/app11083636 ◽

2021 ◽

Vol 11 (8) ◽

pp. 3636

Author(s):

Faria Zarin Subah ◽

Kaushik Deb ◽

Pranab Kumar Dhar ◽

Takeshi Koshiba

Keyword(s):

Autism Spectrum Disorder ◽

Resting State ◽

State Of The Art ◽

Resting State Fmri ◽

Autism Spectrum ◽

Spectrum Disorder ◽

Bootstrap Analysis ◽

Proposed Model ◽

Art Methods ◽

The Mean

Autism spectrum disorder (ASD) is a complex and degenerative neuro-developmental disorder. Most of the existing methods utilize functional magnetic resonance imaging (fMRI) to detect ASD with a very limited dataset which provides high accuracy but results in poor generalization. To overcome this limitation and to enhance the performance of the automated autism diagnosis model, in this paper, we propose an ASD detection model using functional connectivity features of resting-state fMRI data. Our proposed model utilizes two commonly used brain atlases, Craddock 200 (CC200) and Automated Anatomical Labelling (AAL), and two rarely used atlases Bootstrap Analysis of Stable Clusters (BASC) and Power. A deep neural network (DNN) classifier is used to perform the classification task. Simulation results indicate that the proposed model outperforms state-of-the-art methods in terms of accuracy. The mean accuracy of the proposed model was 88%, whereas the mean accuracy of the state-of-the-art methods ranged from 67% to 85%. The sensitivity, F1-score, and area under receiver operating characteristic curve (AUC) score of the proposed model were 90%, 87%, and 96%, respectively. Comparative analysis on various scoring strategies show the superiority of BASC atlas over other aforementioned atlases in classifying ASD and control.

Download Full-text

Sex Differences in Functional Connectivity Between Resting State Brain Networks in Autism Spectrum Disorder

Journal of Autism and Developmental Disorders ◽

10.1007/s10803-021-05191-6 ◽

2021 ◽

Author(s):

Vânia Tavares ◽

Luís Afonso Fernandes ◽

Marília Antunes ◽

Hugo Ferreira ◽

Diana Prata

Keyword(s):

Autism Spectrum Disorder ◽

Sex Differences ◽

Resting State ◽

Brain Connectivity ◽

Autism Spectrum ◽

Spectrum Disorder ◽

Control Networks ◽

Highly Sensitive ◽

Female Specific ◽

Main Effects

AbstractFunctional brain connectivity (FBC) has previously been examined in autism spectrum disorder (ASD) between-resting-state networks (RSNs) using a highly sensitive and reproducible hypothesis-free approach. However, results have been inconsistent and sex differences have only recently been taken into consideration using this approach. We estimated main effects of diagnosis and sex and a diagnosis by sex interaction on between-RSNs FBC in 83 ASD (40 females/43 males) and 85 typically developing controls (TC; 43 females/42 males). We found increased connectivity between the default mode (DM) and (a) the executive control networks in ASD (vs. TC); (b) the cerebellum networks in males (vs. females); and (c) female-specific altered connectivity involving visual, language and basal ganglia (BG) networks in ASD—in suggestive compatibility with ASD cognitive and neuroscientific theories.

Download Full-text

Increased Functional Connectivity Between Subcortical and Cortical Resting-State Networks in Autism Spectrum Disorder

JAMA Psychiatry ◽

10.1001/jamapsychiatry.2015.0101 ◽

2015 ◽

Vol 72 (8) ◽

pp. 767 ◽

Cited By ~ 138

Author(s):

Leonardo Cerliani ◽

Maarten Mennes ◽

Rajat M. Thomas ◽

Adriana Di Martino ◽

Marc Thioux ◽

...

Keyword(s):

Autism Spectrum Disorder ◽

Functional Connectivity ◽

Resting State ◽

Autism Spectrum ◽

Spectrum Disorder ◽

Resting State Networks

Download Full-text

Graph Ricci Curvatures Reveal Disease-related Changes in Autism Spectrum Disorder

10.1101/2021.11.28.470231 ◽

2021 ◽

Author(s):

Pavithra Elumalai ◽

Yasharth Yadav ◽

Nitin Williams ◽

Emil Saucan ◽

Jürgen Jost ◽

...

Keyword(s):

Autism Spectrum Disorder ◽

Resting State ◽

Neurodevelopmental Disorders ◽

Data Exchange ◽

Meta Analysis ◽

Resting State Fmri ◽

Brain Regions ◽

Autism Spectrum ◽

Spectrum Disorder ◽

The Brain

Autism Spectrum Disorder (ASD) is a set of neurodevelopmental disorders that pose a significant global health burden. Measures from graph theory have been used to characterise ASD-related changes in resting-state fMRI functional connectivity networks (FCNs), but recently developed geometry-inspired measures have not been applied so far. In this study, we applied geometry-inspired graph Ricci curvatures to investigate ASD-related changes in resting-state fMRI FCNs. To do this, we applied Forman-Ricci and Ollivier-Ricci curvatures to compare networks of ASD and healthy controls (N = 1112) from the Autism Brain Imaging Data Exchange I (ABIDE-I) dataset. We performed these comparisons at the brain-wide level as well as at the level of individual brain regions, and further, determined the behavioral relevance of region-specific differences with Neurosynth meta-analysis decoding. We found brain-wide ASD-related differences for both Forman-Ricci and Ollivier-Ricci curvatures. For Forman-Ricci curvature, these differences were distributed across 83 of the 200 brain regions studied, and concentrated within the Default Mode, Somatomotor and Ventral Attention Network. Meta-analysis decoding identified the brain regions showing curvature differences as involved in social cognition, memory, language and movement. Notably, comparison with results from previous non-invasive stimulation (TMS/tDCS) experiments revealed that the set of brain regions showing curvature differences overlapped with the set of brain regions whose stimulation resulted in positive cognitive or behavioural outcomes in ASD patients. These results underscore the utility of geometry-inspired graph Ricci curvatures in characterising disease-related changes in ASD, and possibly, other neurodevelopmental disorders.

Download Full-text

Multisite Autism Spectrum Disorder Classification Using Convolutional Neural Network Classifier and Individual Morphological Brain Networks

Frontiers in Neuroscience ◽

10.3389/fnins.2020.629630 ◽

2021 ◽

Vol 14 ◽

Author(s):

Jingjing Gao ◽

Mingren Chen ◽

Yuanyuan Li ◽

Yachun Gao ◽

Yanling Li ◽

...

Keyword(s):

Neural Network ◽

Autism Spectrum Disorder ◽

Convolutional Neural Network ◽

Structural Mri ◽

Autism Spectrum ◽

Brain Network ◽

Spectrum Disorder ◽

Structural Covariance ◽

Almost All ◽

New Framework

Autism spectrum disorder (ASD) is a range of neurodevelopmental disorders with behavioral and cognitive impairment and brings huge burdens to the patients’ families and the society. To accurately identify patients with ASD from typical controls is important for early detection and early intervention. However, almost all the current existing classification methods for ASD based on structural MRI (sMRI) mainly utilize the independent local morphological features and do not consider the covariance patterns of these features between regions. In this study, by combining the convolutional neural network (CNN) and individual structural covariance network, we proposed a new framework to classify ASD patients with sMRI data from the ABIDE consortium. Moreover, gradient-weighted class activation mapping (Grad-CAM) was applied to characterize the weight of features contributing to the classification. The experimental results showed that our proposed method outperforms the currently used methods for classifying ASD patients with the ABIDE data and achieves a high classification accuracy of 71.8% across different sites. Furthermore, the discriminative features were found to be mainly located in the prefrontal cortex and cerebellum, which may be the early biomarkers for the diagnosis of ASD. Our study demonstrated that CNN is an effective tool to build the framework for the diagnosis of ASD with individual structural covariance brain network.

Download Full-text