Dataset Information

Remodeling Pearson's Correlation for Functional Brain Network Estimation and Autism Spectrum Disorder Identification.

ABSTRACT: Functional brain network (FBN) has been becoming an increasingly important way to model the statistical dependence among neural time courses of brain, and provides effective imaging biomarkers for diagnosis of some neurological or psychological disorders. Currently, Pearson's Correlation (PC) is the simplest and most widely-used method in constructing FBNs. Despite its advantages in statistical meaning and calculated performance, the PC tends to result in a FBN with dense connections. Therefore, in practice, the PC-based FBN needs to be sparsified by removing weak (potential noisy) connections. However, such a scheme depends on a hard-threshold without enough flexibility. Different from this traditional strategy, in this paper, we propose a new approach for estimating FBNs by remodeling PC as an optimization problem, which provides a way to incorporate biological/physical priors into the FBNs. In particular, we introduce an L₁-norm regularizer into the optimization model for obtaining a sparse solution. Compared with the hard-threshold scheme, the proposed framework gives an elegant mathematical formulation for sparsifying PC-based networks. More importantly, it provides a platform to encode other biological/physical priors into the PC-based FBNs. To further illustrate the flexibility of the proposed method, we extend the model to a weighted counterpart for learning both sparse and scale-free networks, and then conduct experiments to identify autism spectrum disorders (ASD) from normal controls (NC) based on the constructed FBNs. Consequently, we achieved an 81.52% classification accuracy which outperforms the baseline and state-of-the-art methods.

SUBMITTER: Li W

PROVIDER: S-EPMC5583214 | biostudies-literature | 2017

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Remodeling Pearson's Correlation for Functional Brain Network Estimation and Autism Spectrum Disorder Identification.

Li Weikai W Wang Zhengxia Z Zhang Limei L Qiao Lishan L Shen Dinggang D

Frontiers in neuroinformatics 20170831

Functional brain network (FBN) has been becoming an increasingly important way to model the statistical dependence among neural time courses of brain, and provides effective imaging biomarkers for diagnosis of some neurological or psychological disorders. Currently, Pearson's Correlation (PC) is the simplest and most widely-used method in constructing FBNs. Despite its advantages in statistical meaning and calculated performance, the PC tends to result in a FBN with dense connections. Therefore, ...[more]

PMID: 28912708

Similar Datasets

Project description:Background: Individuals with autism spectrum disorder (ASD) show elevated levels of motor variability that are associated with clinical outcomes. Cortical-cerebellar networks involved in visuomotor control have been implicated in postmortem and anatomical imaging studies of ASD. However, the extent to which these networks show intrinsic functional alterations in patients, and the relationship between intrinsic functional properties of cortical-cerebellar networks and visuomotor impairments in ASD have not yet been clarified. Methods: We examined the amplitude of low-frequency fluctuation (ALFF) of cortical and cerebellar brain regions during resting-state functional MRI (rs-fMRI) in 23 individuals with ASD and 16 typically developing (TD) controls. Regions of interest (ROIs) with ALFF values significantly associated with motor variability were identified for for patients and controls respectively, and their functional connectivity (FC) to each other and to the rest of the brain was examined. Results: For TD controls, greater ALFF in bilateral cerebellar crus I, left superior temporal gyrus, left inferior frontal gyrus, right supramarginal gyrus, and left angular gyrus each were associated with greater visuomotor variability. Greater ALFF in cerebellar lobule VIII was associated with less visuomotor variability. For individuals with ASD, greater ALFF in right calcarine cortex, right middle temporal gyrus (including MT/V5), left Heschl's gyrus, left post-central gyrus, right pre-central gyrus, and left precuneus was related to greater visuomotor variability. Greater ALFF in cerebellar vermis VI was associated with less visuomotor variability. Individuals with ASD and TD controls did not show differences in ALFF for any of these ROIs. Individuals with ASD showed greater posterior cerebellar connectivity with occipital and parietal cortices relative to TD controls, and reduced FC within cerebellum and between lateral cerebellum and pre-frontal and other regions of association cortex. Conclusion: Together, these findings suggest that increased resting oscillations within visuomotor networks in ASD are associated with more severe deficits in controlling variability during precision visuomotor behavior. Differences between individuals with ASD and TD controls in the topography of networks showing relationships to visuomotor behavior suggest atypical patterns of cerebellar-cortical specialization and connectivity in ASD that underlies previously documented visuomotor deficits.

Dataset Information

Remodeling Pearson's Correlation for Functional Brain Network Estimation and Autism Spectrum Disorder Identification.

Publications

Remodeling Pearson's Correlation for Functional Brain Network Estimation and Autism Spectrum Disorder Identification.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets