Dynamic functional connectivity analysis reveals decreased variability of the default‐mode network in developing autistic brain

Abstract

Accumulating neuroimaging evidence suggests that abnormal functional connectivity of the default mode network (DMN) contributes to the social‐cognitive deficits of autism spectrum disorder (ASD). Although most previous studies relied on conventional functional connectivity methods, which assume that connectivity patterns remain constant over time, understanding the temporal dynamics of functional connectivity during rest may provide new insights into the dysfunction of the DMN in ASD. In this work, dynamic functional connectivity analysis based on sliding time window correlation was applied to the resting‐state functional magnetic resonance imaging data of 28 young children with ASD (age range: 3–7 years) and 29 matched typically developing controls (TD group). In addition, k‐means cluster analysis was performed to identify distinct temporal states based on the spatial similarity of each functional connectivity pattern. Compared with the TD group, young children with ASD showed decreased dynamic functional connectivity variance between the posterior cingulate cortex (PCC) and the right precentral gyrus, which is negatively correlated with social motivation and social relating. Cluster analysis revealed significant differences in functional connectivity patterns between the ASD and TD groups in discrete temporal states. Our findings reveal that atypical dynamic interactions between the PCC and sensorimotor cortex are associated with social deficits in ASD. Results also highlight the critical role of PCC in the social‐cognitive deficits of ASD and support the concept that understanding the dynamic neural interactions among brain regions can provide insights into functional abnormalities in ASD. Autism Research 2018, 11: 1479–1493. © 2018 International Society for Autism Research, Wiley Periodicals, Inc. Lay Summary Social cognitive dysfunction in autism spectrum disorder (ASD) is associated with dysfunction of the default mode network (DMN), a set of brain areas involved in various domains of social processing. We found that decreases in the dynamic functional connectivity variance between the posterior cingulate cortex and the sensorimotor cortex are associated with deficits in social motivation and social relating in young children with ASD. This result suggests that aberrations in the DMN and its dynamic interactions with other networks contribute to atypical integration of information with respect to self and others.

Altered inter‐ and intrahemispheric functional connectivity dynamics in autistic children

“…Previous dynamic functional connectivity studies demonstrated that too short window length can increase the risk of introducing artifacts in dynamic functional connectivity analysis and reducing reliability of functional connectivity, and too long window lengths can obscure the temporal variations of dynamic functional connectivity (Preti, Bolton, & Ville, ). Thus 50 TRs was selected as the window length in the current study to balance the specificity and sensitivity for dynamic functional connectivity calculation (He et al, ; Li, Liao, et al, ). To verify the robustness of the sliding‐window analysis, we replicated our findings with different window lengths (30 TRs and 80 TRs) and shifting step (1TR) (Figures S1 and S2, Supporting Information).…”

“…Capturing the functional connectivity dynamics has been validated to provide new insights into the brain networks of neuropsychiatric disorders, including epilepsy (Li et al, ), depression (Liao et al, ) and schizophrenia (Damaraju et al, ). Investigations of dynamic functional connectivity in ASD have revealed aberrant dynamics of functional connectivity (Chen, Nomi, Uddin, Duan, & Chen, ; Guo et al, ; He et al, ). For example, enhanced temporal variability of intrinsic functional connectivity was observed in individuals with ASD, implying increased intra‐individual variance of brain networks across time (Falahpour et al, ).…”

Dynamic Functional Connectivity Reveals Abnormal Variability and Hyper‐connected Pattern in Autism Spectrum Disorder

“…Then, rsfMRI data were preprocessed by Data Processing & Analysis for Brain Imaging (DPABI V3.0, http://rfmri.org/) [Yan, Wang, Zuo, & Zang, ], an open‐source package based on Statistical Parametric Mapping (SPM8, https://www.fil.ion.ucl.ac.uk/spm/) and MATLAB (MathWorks). For every subject, the steps of image preprocessing were as follows: (a) the first five volumes were discarded to allow magnetization stabilization; (b) time‐slicing and realignment were performed; (c) functional and structural images were manually reoriented; (d) structural images were co‐registered to functional images and segmented into gray matter, white matter, and cerebrospinal fluid; (e) nuisance covariates were regressed (including Friston 24 head motion parameters [Friston, Williams, Howard, Frackowiak, & Turner, ] and white matter and cerebrospinal fluid signals); (f) functional images were normalized into Montreal Neurological Institute standard space by Diffeomorphic Anatomical Registration Through Exponentiated Lie algebra [Ashburner, ] and resampled to 3.0 × 3.0 × 3.0 mm 3 ; (g) spatial smoothing was performed (Gaussian kernel of 6 mm FWHM); (h) Filtering (0.01–0.08 Hz) was applied to reduce the effects of low‐frequency drifts and high‐frequency aliasing; and (i) image volumes with FD >0.2 mm were scrubbed to reduce the effect of head motion using spline interpolation [He et al, ; Xin et al, ].…”

“…Nevertheless, DFC has not been widely used to assess ASD, and most investigations paid more attention to clustering results [de Lacy, Doherty, King, Rachakonda, & Calhoun, 2017;Guo et al, 2018;He et al, 2018;Mash et al, 2019;Rashid et al, 2018;Yao et al, 2016] instead of metrics such as the SD Falahpour et al, 2016;He et al, 2018]. In our work, we give more interest in these straightforward characteristics, which could provide more insight into neural mechanisms underlying ASD in terms of temporal variability.…”

Partially impaired functional connectivity states between right anterior insula and default mode network in autism spectrum disorder

“…Similar dFC studies in epilepsy revealed state‐specific impairments of functional connectivity patterns in participants with epilepsy compared with healthy controls (Liao, Zhang, et al, ; Liu et al, ). Individuals with ASD were also confirmed to exhibit abnormal temporal variability in functional architecture at rest (Falahpour et al, ; He et al, ; Zhang et al, ). Several studies have examined functional connectivity profiles of the rAI at rest in ASD (Abbott et al, ; Ebisch et al, ; von dem Hagen et al, ).…”