Abstract
Several previous studies have reported atypicality in resting-state functional connectivity (FC) in autism spectrum disorder (ASD), yet the relatively small effect sizes prevent us from using these characteristics for diagnostic purposes. Here, canonical correlation analysis (CCA) and hierarchical clustering were used to partition the high-functioning ASD group (i.e., the ASD discovery group) into subgroups. A support vector machine (SVM) model was trained through the 10-fold strategy to predict Autism Diagnostic Observation Schedule (ADOS) scores within the ASD discovery group (r = 0.30, P < 0.001, n = 260), which was further validated in an independent sample (i.e., the ASD validation group) (r = 0.35, P = 0.031, n = 29). The neuroimage-based partition derived two subgroups representing severe versus mild autistic patients. We identified FCs that show graded changes in strength from ASD-severe, through ASD-mild, to controls, while the same pattern cannot be observed in partitions based on ADOS score. We also identified FCs that are specific for ASD-mild, similar to a partition based on ADOS score. The current study provided multiple pieces of evidence with replication to show that resting-state functional magnetic resonance imaging (rsfMRI) FCs could serve as neural biomarkers in partitioning high-functioning autistic individuals based on their symptom severity and showing advantages over traditional partition based on ADOS score. Our results also indicate a compensatory role for a frontocortical network in patients with mild ASD, indicating potential targets for future clinical treatments.
Highlights
Autism spectrum disorder (ASD) is a neurodevelopmental condition characterized by qualitative impairment in social communication, as well as restricted and repetitive behaviors (American Psychiatric Association, 2013), and affects approximately 1% of children globally (Kim et al, 2011; Baio et al, 2018)
A hierarchical clustering based on the three components partitioned the autism spectrum disorder (ASD) discovery individuals into two subgroups that were further found with significant differences on their Autism Diagnostic Observation Schedule (ADOS) total scores (ASD-severe subgroup: mean = 14.07, SD = 3.27, n = 91; ASD-mild subgroup: mean = 11.07, SD = 2.45, n = 169; Cohen’s D = 1.26, t = 9.23, Ppermutation = 0.009; the permutation process includes the functional connectivity (FC) selection process, canonical correlation analysis (CCA), and the hierarchical clustering and properly adjusted for any possible overfittings; see section “Materials and Methods”; Figure 2C), which confirmed the existence of two cluster
Through a newly developed statistical approach combining CCA and hierarchical clustering to identify candidate neural features that were further trained and independently validated with a machine learning model, our results demonstrated that resting-state functional magnetic resonance imaging (rsfMRI) FC-based stratification of high-functioning ASD patients was effective in differentiating severe from mild ASD patients, showing excellent consistency with traditional behavior-based diagnostic segregations at varied cutoffs of the ADOS scores
Summary
Autism spectrum disorder (ASD) is a neurodevelopmental condition characterized by qualitative impairment in social communication, as well as restricted and repetitive behaviors (American Psychiatric Association, 2013), and affects approximately 1% of children globally (Kim et al, 2011; Baio et al, 2018). No previous MRI study has found effect sizes large enough to indicate that brain structure or function could be used as a diagnostic marker. This has prompted a shift to focus on the identification of stratification biomarkers to parse this heterogeneous condition into more homogeneous subgroups (Loth et al, 2016). Few studies have evaluated the agreement between biomarker-based stratification of ASD patients and differences in clinical symptom profile or severity
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have