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

  • Author Footnotes
    1 KS and SR contributed equally to this work.
    Kaustubh Supekar
    Address correspondence to Kaustubh Supekar, Ph.D.
    1 KS and SR contributed equally to this work.
    Department of Psychiatry and Behavioral Sciences, Stanford University School of Medicine, Stanford, California
    Search for articles by this author
  • Author Footnotes
    1 KS and SR contributed equally to this work.
    Srikanth Ryali
    1 KS and SR contributed equally to this work.
    Department of Psychiatry and Behavioral Sciences, Stanford University School of Medicine, Stanford, California
    Search for articles by this author
  • Rui Yuan
    Department of Psychiatry and Behavioral Sciences, Stanford University School of Medicine, Stanford, California
    Search for articles by this author
  • Devinder Kumar
    Department of Psychiatry and Behavioral Sciences, Stanford University School of Medicine, Stanford, California
    Search for articles by this author
  • Carlo de los Angeles
    Department of Psychiatry and Behavioral Sciences, Stanford University School of Medicine, Stanford, California
    Search for articles by this author
  • Vinod Menon
    Vinod Menon, Ph.D.
    Department of Psychiatry and Behavioral Sciences, Stanford University School of Medicine, Stanford, California

    Department of Neurology & Neurological Sciences, Stanford University School of Medicine, Stanford, California

    Stanford Neurosciences Institute, Stanford University School of Medicine, Stanford, California
    Search for articles by this author
  • Author Footnotes
    1 KS and SR contributed equally to this work.
Published:February 15, 2022DOI:



      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.


      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.


      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.


      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.


      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 to Biological Psychiatry
      Already a print subscriber? Claim online access
      Already an online subscriber? Sign in
      Institutional Access: Sign in to ScienceDirect


        • Kanner L.
        Autistic disturbances of affective contact.
        Nervous Child. 1943; 2: 217-250
        • Klin A.
        Young autistic children’s listening preferences in regard to speech: A possible characterization of the symptom of social withdrawal.
        J Autism Dev Disord. 1991; 21: 29-42
        • Maenner M.J.
        • Shaw K.A.
        • Baio J.
        • Washington A.
        • Patrick M.
        • DiRienzo M.
        • et al.
        Prevalence of autism spectrum disorder among children aged 8 years—Autism and Developmental Disabilities Monitoring Network, 11 sites, United States, 2016 [published correction appears in MMWR Surveill Summ 2020; 69:503].
        MMWR Surveill Summ. 2020; 69: 1-12
        • Lord C.
        • Brugha T.S.
        • Charman T.
        • Cusack J.
        • Dumas G.
        • Frazier T.
        • et al.
        Autism spectrum disorder.
        Nat Rev Dis Primers. 2020; 6: 5
        • Lombardo M.V.
        • Lai M.C.
        • Baron-Cohen S.
        Big data approaches to decomposing heterogeneity across the autism spectrum.
        Mol Psychiatry. 2019; 24: 1435-1450
        • Liu W.
        • Wang Z.
        • Liu X.
        • Zeng N.
        • Liu Y.
        • Alsaadi F.E.
        A survey of deep neural network architectures and their applications.
        Neurocomputing. 2017; 234: 11-26
        • Durstewitz D.
        • Koppe G.
        • Meyer-Lindenberg A.
        Deep neural networks in psychiatry.
        Mol Psychiatry. 2019; 24: 1583-1598
        • Aghdam M.A.
        • Sharifi A.
        • Pedram M.M.
        Diagnosis of autism spectrum disorders in young children based on resting-state functional magnetic resonance imaging data using convolutional neural networks.
        J Digit Imaging. 2019; 32: 899-918
        • Dvornek N.C.
        • Ventola P.
        • Duncan J.S.
        Combining phenotypic and resting-state fMRI data for autism classification with recurrent neural networks.
        Proc IEEE Int Symp Biomed Imaging. 2018; 2018: 725-728
        • Dvornek N.C.
        • Ventola P.
        • Pelphrey K.A.
        • Duncan J.S.
        Identifying autism from resting-state fMRI using long short-term memory networks.
        Mach Learn Med Imaging. 2017; 10541: 362-370
        • Quaak M.
        • van de Mortel L.
        • Thomas R.M.
        • van Wingen G.
        Deep learning applications for the classification of psychiatric disorders using neuroimaging data: Systematic review and meta-analysis.
        Neuroimage Clin. 2021; 30: 102584
        • Heinsfeld A.S.
        • Franco A.R.
        • Craddock R.C.
        • Buchweitz A.
        • Meneguzzi F.
        Identification of autism spectrum disorder using deep learning and the ABIDE dataset.
        Neuroimage Clin. 2017; 17: 16-23
        • Sherkatghanad Z.
        • Akhondzadeh M.
        • Salari S.
        • Zomorodi-Moghadam M.
        • Abdar M.
        • Acharya U.R.
        • et al.
        Automated detection of autism spectrum disorder using a convolutional neural network.
        Front Neurosci. 2020; 13: 1325
        • Rakić M.
        • Cabezas M.
        • Kushibar K.
        • Oliver A.
        • Lladó X.
        Improving the detection of autism spectrum disorder by combining structural and functional MRI information.
        Neuroimage Clin. 2020; 25: 102181
        • Zhuang J.
        • Dvornek N.C.
        • Li X.
        • Ventola P.
        • Duncan J.S.
        Invertible network for classification and biomarker selection for ASD.
        Med Image Comput Comput Assist Interv. 2019; 11766: 700-708
        • Rathore A.
        • Palande S.
        • Anderson J.S.
        • Zielinski B.A.
        • Fletcher P.T.
        • Wang B.
        Autism classification using topological features and deep learning: A cautionary tale.
        Med Image Comput Comput Assist Interv. 2019; 11766: 736-744
        • Ryali S.
        • Supekar K.
        • Chen T.
        • Kochalka J.
        • Cai W.
        • Nicholas J.
        • et al.
        Temporal dynamics and developmental maturation of salience, default and central-executive network interactions revealed by variational Bayes hidden Markov modeling.
        PLoS Comput Biol. 2016; 12e1005138
        • Taghia J.
        • Cai W.
        • Ryali S.
        • Kochalka J.
        • Nicholas J.
        • Chen T.
        • Menon V.
        Uncovering hidden brain state dynamics that regulate performance and decision-making during cognition.
        Nat Commun. 2018; 9: 2505
        • Ryali S.
        • Supekar K.
        • Chen T.
        • Menon V.
        Multivariate dynamical systems models for estimating causal interactions in fMRI.
        Neuroimage. 2011; 54: 807-823
        • Koppe G.
        • Meyer-Lindenberg A.
        • Durstewitz D.
        Deep learning for small and big data in psychiatry.
        Neuropsychopharmacology. 2021; 46: 176-190
        • Calhoun V.D.
        • Miller R.
        • Pearlson G.
        • Adalı T.
        The chronnectome: Time-varying connectivity networks as the next frontier in fMRI data discovery.
        Neuron. 2014; 84: 262-274
        • Cai W.
        • Chen T.
        • Szegletes L.
        • Supekar K.
        • Menon V.
        Aberrant time-varying cross-network interactions in children with attention-deficit/hyperactivity disorder and the relation to attention deficits.
        Biol Psychiatry Cogn Neurosci Neuroimaging. 2018; 3: 263-273
        • Supekar K.
        • Cai W.
        • Krishnadas R.
        • Palaniyappan L.
        • Menon V.
        Dysregulated brain dynamics in a triple-network saliency model of schizophrenia and its relation to psychosis.
        Biol Psychiatry. 2019; 85: 60-69
        • Bolton T.A.W.
        • Morgenroth E.
        • Preti M.G.
        • Van De Ville D.
        Tapping into multi-faceted human behavior and psychopathology using fMRI brain dynamics.
        Trends Neurosci. 2020; 43: 667-680
        • Preti M.G.
        • Bolton T.A.
        • Van De Ville D.
        The dynamic functional connectome: State-of-the-art and perspectives.
        Neuroimage. 2017; 160: 41-54
        • Christoff K.
        • Irving Z.C.
        • Fox K.C.R.
        • Spreng R.N.
        • Andrews-Hanna J.R.
        Mind-wandering as spontaneous thought: A dynamic framework.
        Nat Rev Neurosci. 2016; 17: 718-731
        • Di Martino A.
        • O’Connor D.
        • Chen B.
        • Alaerts K.
        • Anderson J.S.
        • Assaf M.
        • et al.
        Enhancing studies of the connectome in autism using the autism brain imaging data exchange II.
        Sci Data. 2017; 4: 170010
        • Di Martino A.
        • Yan C.G.
        • Li Q.
        • Denio E.
        • Castellanos F.X.
        • Alaerts K.
        • et al.
        The autism brain imaging data exchange: Towards a large-scale evaluation of the intrinsic brain architecture in autism.
        Mol Psychiatry. 2014; 19: 659-667
        • Padmanabhan A.
        • Lynch C.J.
        • Schaer M.
        • Menon V.
        The default mode network in autism.
        Biol Psychiatry Cogn Neurosci Neuroimaging. 2017; 2: 476-486
        • Abrams D.A.
        • Padmanabhan A.
        • Chen T.
        • Odriozola P.
        • Baker A.E.
        • Kochalka J.
        • et al.
        Impaired voice processing in reward and salience circuits predicts social communication in children with autism.
        Elife. 2019; 8e39906
        • Abrams D.A.
        • Lynch C.J.
        • Cheng K.M.
        • Phillips J.
        • Supekar K.
        • Ryali S.
        • et al.
        Underconnectivity between voice-selective cortex and reward circuitry in children with autism.
        Proc Natl Acad Sci U S A. 2013; 110: 12060-12065
        • Lynch C.J.
        • Uddin L.Q.
        • Supekar K.
        • Khouzam A.
        • Phillips J.
        • Menon V.
        Default mode network in childhood autism: Posteromedial cortex heterogeneity and relationship with social deficits.
        Biol Psychiatry. 2013; 74: 212-219
        • Davatzikos C.
        Machine learning in neuroimaging: Progress and challenges.
        Neuroimage. 2019; 197: 652-656
        • Fan L.
        • Li H.
        • Zhuo J.
        • Zhang Y.
        • Wang J.
        • Chen L.
        • et al.
        The human Brainnetome atlas: A new brain atlas based on connectional architecture.
        Cereb Cortex. 2016; 26: 3508-3526
        • Lundberg S.M.
        • Lee S.I.
        A unified approach to interpreting model predictions.
        in: von Luxburg U. Guyon I. Bengio S. Wallach H. Fergus R. NIPS ’17: Proceedings of the 31st International Conference on Neural Information Processing Systems. Curran Associates, Red Hook, NY2017: 4768-4777
        • Simonyan K.
        • Vedaldi A.
        • Zisserman A.
        Deep inside convolutional networks: Visualising image classification models and saliency maps.
        arXiv. 2013;
        • Springenberg J.T.
        • Dosovitskiy A.
        • Brox T.
        • Riedmiller M.
        Striving for simplicity: The all convolutional net.
        arXiv. 2014;
        • Selvaraju R.R.
        • Cogswell M.
        • Das A.
        • Vedantam R.
        • Parikh D.
        • Batra D.
        Grad-CAM: Visual explanations from deep networks via gradient-based localization.
        2017 IEEE International Conference on Computer Vision (ICCV). 2017; : 618-626
        • Sundararajan M.
        • Taly A.
        • Yan Q.
        Axiomatic attribution for deep networks.
        in: Precup D. Teh Y.W. ICML ‘17: Proceedings of the 34th International Conference on Machine Learning-Volume 70., 2017: 3319-3328
        • Abrams D.A.
        • Kochalka J.
        • Bhide S.
        • Ryali S.
        • Menon V.
        Intrinsic functional architecture of the human speech processing network.
        Cortex. 2020; 129: 41-56
        • Belin P.
        • Zatorre R.J.
        • Lafaille P.
        • Ahad P.
        • Pike B.
        Voice-selective areas in human auditory cortex.
        Nature. 2000; 403: 309-312
        • Kanwisher N.
        • McDermott J.
        • Chun M.M.
        The fusiform face area: A module in human extrastriate cortex specialized for face perception.
        J Neurosci. 1997; 17: 4302-4311
        • Gervais H.
        • Belin P.
        • Boddaert N.
        • Leboyer M.
        • Coez A.
        • Sfaello I.
        • et al.
        Abnormal cortical voice processing in autism.
        Nat Neurosci. 2004; 7: 801-802
        • Hus V.
        • Lord C.
        The autism diagnostic observation schedule, module 4: Revised algorithm and standardized severity scores.
        J Autism Dev Disord. 2014; 44: 1996-2012
        • Gotham K.
        • Pickles A.
        • Lord C.
        Standardizing ADOS scores for a measure of severity in autism spectrum disorders.
        J Autism Dev Disord. 2009; 39: 693-705
        • Di Martino A.
        • Zuo X.N.
        • Kelly C.
        • Grzadzinski R.
        • Mennes M.
        • Schvarcz A.
        • et al.
        Shared and distinct intrinsic functional network centrality in autism and attention-deficit/hyperactivity disorder.
        Biol Psychiatry. 2013; 74: 623-632
        • Schilbach L.
        • Eickhoff S.B.
        • Rotarska-Jagiela A.
        • Fink G.R.
        • Vogeley K.
        Minds at rest? Social cognition as the default mode of cognizing and its putative relationship to the “default system” of the brain.
        Conscious Cogn. 2008; 17: 457-467
        • Mars R.B.
        • Neubert F.X.
        • Noonan M.P.
        • Sallet J.
        • Toni I.
        • Rushworth M.F.S.
        On the relationship between the “default mode network” and the “social brain.”.
        Front Hum Neurosci. 2012; 6: 189
        • Spreng R.N.
        • Mar R.A.
        • Kim A.S.N.
        The common neural basis of autobiographical memory, prospection, navigation, theory of mind, and the default mode: A quantitative meta-analysis.
        J Cogn Neurosci. 2009; 21: 489-510
        • Laird A.R.
        • Fox P.M.
        • Eickhoff S.B.
        • Turner J.A.
        • Ray K.L.
        • McKay D.R.
        • et al.
        Behavioral interpretations of intrinsic connectivity networks.
        J Cogn Neurosci. 2011; 23: 4022-4037
        • Smith S.M.
        • Fox P.T.
        • Miller K.L.
        • Glahn D.C.
        • Fox P.M.
        • Mackay C.E.
        • et al.
        Correspondence of the brain’s functional architecture during activation and rest.
        Proc Natl Acad Sci U S A. 2009; 106: 13040-13045
        • Haar S.
        • Berman S.
        • Behrmann M.
        • Dinstein I.
        Anatomical abnormalities in autism?.
        Cereb Cortex. 2016; 26: 1440-1452
        • Valk S.L.
        • Di Martino A.
        • Milham M.P.
        • Bernhardt B.C.
        Multicenter mapping of structural network alterations in autism.
        Hum Brain Mapp. 2015; 36: 2364-2373
        • Uddin L.Q.
        • Menon V.
        • Young C.B.
        • Ryali S.
        • Chen T.
        • Khouzam A.
        • et al.
        Multivariate searchlight classification of structural magnetic resonance imaging in children and adolescents with autism.
        Biol Psychiatry. 2011; 70: 833-841
        • Moseley R.L.
        • Ypma R.J.F.
        • Holt R.J.
        • Floris D.
        • Chura L.R.
        • Spencer M.D.
        • et al.
        Whole-brain functional hypoconnectivity as an endophenotype of autism in adolescents.
        Neuroimage Clin. 2015; 9: 140-152
        • Glerean E.
        • Pan R.K.
        • Salmi J.
        • Kujala R.
        • Lahnakoski J.M.
        • Roine U.
        • et al.
        Reorganization of functionally connected brain subnetworks in high-functioning autism.
        Hum Brain Mapp. 2016; 37: 1066-1079
        • Yerys B.E.
        • Gordon E.M.
        • Abrams D.N.
        • Satterthwaite T.D.
        • Weinblatt R.
        • Jankowski K.F.
        • et al.
        Default mode network segregation and social deficits in autism spectrum disorder: Evidence from non-medicated children.
        Neuroimage Clin. 2015; 9: 223-232
        • Kuhl P.K.
        • Conboy B.T.
        • Padden D.
        • Nelson T.
        • Pruitt J.
        Early speech perception and later language development: Implications for the “critical period.”.
        Lang Learn Dev. 2005; 1: 237-264
        • Christophe A.
        • Dupoux E.
        • Bertoncini J.
        • Mehler J.
        Do infants perceive word boundaries? An empirical study of the bootstrapping of lexical acquisition.
        J Acoust Soc Am. 1994; 95: 1570-1580
        • DeCasper A.J.
        • Fifer W.P.
        Of human bonding: Newborns prefer their mothers’ voices.
        Science. 1980; 208: 1174-1176
        • Harstad L.
        • Baum C.
        • Yatchmink Y.
        Early warning signs of autism spectrum disorder. National Center on Birth Defects and Developmental Disabilities, Centers for Disease Control and Prevention.
        (Available at:)
        • Demetriou E.A.
        • Lampit A.
        • Quintana D.S.
        • Naismith S.L.
        • Song Y.J.C.
        • Pye J.E.
        • et al.
        Autism spectrum disorders: A meta-analysis of executive function.
        Mol Psychiatry. 2018; 23: 1198-1204
        • Hill E.L.
        Executive dysfunction in autism.
        Trends Cogn Sci. 2004; 8: 26-32
        • Lai C.L.E.
        • Lau Z.
        • Lui S.S.Y.
        • Lok E.
        • Tam V.
        • Chan Q.
        • et al.
        Meta-analysis of neuropsychological measures of executive functioning in children and adolescents with high-functioning autism spectrum disorder.
        Autism Res. 2017; 10: 911-939
        • Oblak A.L.
        • Rosene D.L.
        • Kemper T.L.
        • Bauman M.L.
        • Blatt G.J.
        Altered posterior cingulate cortical cyctoarchitecture, but normal density of neurons and interneurons in the posterior cingulate cortex and fusiform gyrus in autism.
        Autism Res. 2011; 4: 200-211

      Linked Article

      • Autism Spectrum Disorder: Time to Notice the Individuals More Than the Group
        Biological PsychiatryVol. 92Issue 8
        • Preview
          Autism spectrum disorder (ASD) is notoriously heterogeneous in its clinical presentation. Moreover, its definition has been revised many times since its first discovery, leading to inconsistent prevalence rates and characteristics among affected individuals across nations and sociodemographic groups depending on the awareness (1). ASD is diagnosed based on a spectrum of behavioral symptoms broadly consisting of social communication deficits and restricted/repetitive behaviors. Many machine learning classification models have been trained using imaging features to discover biological underpinnings but with only limited success (2).
        • Full-Text
        • PDF