- •The proposed method improves the identification of autism on resting-state functional magnetic imaging data.
- •The most discriminative feature subset about functional connectivity in the brain of autism was found.
- •The classification accuracy obtained is better than the recent similar studies.
- •The results can be used as an important reference for autism diagnoses.
Considering the unsatisfactory classification accuracy of autism due to unsuitable features selected in current studies, a functional connectivity (FC)-based algorithm for classifying autism and control using support vector machine-recursive feature elimination (SVM-RFE) is proposed in this paper. The goal is to find the optimal features based on FC and improve the classification accuracy on a large sample of data. We chose 35 regions of interest based on the social motivation hypothesis to construct the FC matrix and searched for informative features in the complex high-dimensional FC dataset by the SVM-RFE with a stratified-4-fold cross-validation strategy. The selected features were then entered into an SVM with a Gaussian kernel for classification. A total of 255 subjects with autism and 276 subjects with typical development from 10 sites were involved in the study. For the data of global sites, the proposed classification algorithm could identify the two groups with an accuracy of 90.60% (sensitivity 90.62%, specificity 90.58%). For the leave-one-site-out test, the proposed algorithm achieved a classification accuracy of 75.00%–95.23% for data from different sites. These promising results demonstrate that the proposed classification algorithm performs better than those in recent similar studies in that the importance of features can be measured accurately and only the most discriminative feature subset is selected.
To read this article in full you will need to make a payment
Purchase one-time access:Academic & Personal: 24 hour online accessCorporate R&D Professionals: 24 hour online access
One-time access price info
- For academic or personal research use, select 'Academic and Personal'
- For corporate R&D use, select 'Corporate R&D Professionals'
Subscribe:Subscribe to Physica Medica: European Journal of Medical Physics
Already a print subscriber? Claim online access
Already an online subscriber? Sign in
Register: Create an account
Institutional Access: Sign in to ScienceDirect
- Autism and abnormal development of brain connectivity.J Neurosci. 2004; 24: 9228-9231https://doi.org/10.1523/JNEUROSCI.3340-04.2004
- Cortical activation and synchronization during sentence comprehension in high-functioning autism: evidence of underconnectivity.Brain. 2004; 127: 1811-1821https://doi.org/10.1093/brain/awh199
- Autism spectrum disorders: developmental disconnection syndromes.Curr Opin Neurobiol. 2007; 17: 103-111https://doi.org/10.1016/j.conb.2007.01.009
- Radial cytoarchitecture and patterns of cortical connectivity in autism.Philos Trans R Soc Lond B Biol Sci. 2009; 364: 1433-1436https://doi.org/10.1098/rstb.2008.0331
- Underconnected, but how? A survey of functional connectivity MRI studies in autism spectrum disorders.Cereb Cortex. 2011; 21: 2233-2243https://doi.org/10.1093/cercor/bhq296
- The autism brain imaging data exchange: towards a large-scale evaluation of the intrinsic brain architecture in autism.Mol Psychiatry. 2014; 19: 659-667https://doi.org/10.1038/mp.2013.78
- Functional connectivity magnetic resonance imaging classification of autism.Brain. 2011; 134https://doi.org/10.1093/brain/awr263
- Differential deactivation during mentalizing and classification of autism based on default mode network connectivity.PLoS One. 2012; 7e50064https://doi.org/10.1371/journal.pone.0050064
- Salience network–based classification and prediction of symptom severity in children with autism.JAMA Psychiatry. 2013; 70: 869-879https://doi.org/10.1001/jamapsychiatry.2013.104
- Multisite functional connectivity MRI classification of autism: ABIDE results.Front Hum Neurosci. 2013; 7: 599https://doi.org/10.3389/fnhum.2013.00599
- Functional connectivity classification of autism identifies highly predictive brain features but falls short of biomarker.Neuroimage Clin. 2015; 7: 359-366https://doi.org/10.1016/j.nicl.2014.12.013
- Deriving reproducible biomarkers from multi-site resting-state data: an Autism-based example.Neuro Image. 2017; 147: 736-745https://doi.org/10.1016/j.neuroimage.2016.10.045
- Identification of autism spectrum disorder using deep learning and the ABIDE dataset.Neuroimage Clin. 2018; 17: 16-23https://doi.org/10.1016/j.nicl.2017.08.017
- Resting state functional magnetic resonance imaging and neural network classified autism and control.Cortex. 2015; 63: 55-67https://doi.org/10.1016/j.cortex.2014.08.011
- Gene selection for cancer classification using support vector machines.Mach Learn. 2002; 46: 389-422https://doi.org/10.1023/a:1012487302797
- Selecting feature subsets based on SVM-RFE and the overlapping ratio with applications in bioinformatics.Molecules. 2017; 23https://doi.org/10.3390/molecules23010052
- Multivariate classification of smokers and nonsmokers using SVM-RFE on structural MRI images.Hum Brain Mapp. 2015; 36: 4869-4879https://doi.org/10.1002/hbm.22956
- Evaluation of the social motivation hypothesis of autism: a systematic review and meta-analysis.JAMA Psychiatry. 2018; 75: 797-808https://doi.org/10.1001/jamapsychiatry.2018.1100
- DPARSF: A MATLAB toolbox for “pipeline” data analysis of resting-state fMRI.Front Syst Neurosci. 2010; 4: 13https://doi.org/10.3389/fnsys.2010.00013
- DPABI: data processing & analysis for (resting-state) brain imaging.Neuroinformatics. 2016; 14: 339-351https://doi.org/10.1007/s12021-016-9299-4
- The impact of global signal regression on resting state correlations: are anti-correlated networks introduced?.Neuroimage. 2009; 44: 893-905https://doi.org/10.1016/j.neuroimage.2008.09.036
- Correlations and anticorrelations in resting-state functional connectivity MRI: a quantitative comparison of preprocessing strategies.Neuroimage. 2009; 47: 1408-1416https://doi.org/10.1016/j.neuroimage.2009.05.005
- Improved optimization for the robust and accurate linear registration and motion correction of brain images.Neuroimage. 2002; 17: 825-841https://doi.org/10.1006/nimg.2002.1132
- Scikit-learn: machine learning in python.J Mach Learn Res. 2013; 12: 2825-2830https://doi.org/10.1524/auto.2011.0951
- Multivariate classification of autism spectrum disorder using frequency-specific resting-state functional connectivity-A multicenter study.Prog Neuro Psychoph. 2016; 64: 1-9https://doi.org/10.1016/j.pnpbp.2015.06.014
Published online: August 22, 2019
Accepted: August 13, 2019
Received in revised form: August 8, 2019
Received: April 27, 2019
© 2019 Associazione Italiana di Fisica Medica. Published by Elsevier Ltd. All rights reserved.