Enhancing the representation of functional connectivity networks by fusing multi‐view information for autism spectrum disorder diagnosis

Abstract

Functional connectivity network provides novel insights on how distributed brain regions are functionally integrated, and its deviations from healthy brain have recently been employed to identify biomarkers for neuropsychiatric disorders. However, most of brain network analysis methods utilized features extracted only from one functional connectivity network for brain disease detection and cannot provide a comprehensive representation on the subtle disruptions of brain functional organization induced by neuropsychiatric disorders. Inspired by the principles of multi‐view learning which utilizes information from multiple views to enhance object representation, we propose a novel multiple network based framework to enhance the representation of functional connectivity networks by fusing the common and complementary information conveyed in multiple networks. Specifically, four functional connectivity networks corresponding to the four adjacent values of regularization parameter are generated via a sparse regression model with group constraint ( l2,1‐norm), to enhance the common intrinsic topological structure and limit the error rate caused by different views. To obtain a set of more meaningful and discriminative features, we propose using a modified version of weighted clustering coefficients to quantify the subtle differences of each group‐sparse network at local level. We then linearly fuse the selected features from each individual network via a multi‐kernel support vector machine for autism spectrum disorder (ASD) diagnosis. The proposed framework achieves an accuracy of 79.35%, outperforming all the compared single network methods for at least 7% improvement. Moreover, compared with other multiple network methods, our method also achieves the best performance, that is, with at least 11% improvement in accuracy.

Diagnosis of Autism Spectrum Disorder Using Central-Moment Features From Low- and High-Order Dynamic Resting-State Functional Connectivity Networks

“…(2) Most selected brain regions are associated with emotional expression, language understanding, and motion coordination, such as precentral gyrus, middle frontal gyrus, middle cingulate gyrus, posterior cingulate gyrus, amygdala, angular gyrus, and others. These observations are consistent with previous studies (Qiu et al, 2010;Ecker et al, 2015;Ha et al, 2015b;Huang et al, 2018). For example, we found that SFGmed L (Andrews-Hanna et al, 2014), ANG R (Andrews-Hanna et al, 2014), PCUN L (Urbain et al, 2015), CAL L (Perkins et al, 2015), FFG R (Urbain et al, 2016), INS L (Leung et al, 2015; Urbain et al, 2016) contributed more to ASD identification, which is in line with the recent finding reported in the existing literatures.…”

“…Recently, resting-state fMRI (rs-fMRI) uses bloodoxygenation-level-dependent (BOLD) signals to probe brain activity, which has shown great potential in exploring the in vivo neuronal underpinnings of ASD (Fornito et al, 2015;Liu et al, 2016;Huang et al, 2018;Zhao et al, 2018). Since BOLD signals are sensitive to the spontaneous and intrinsic neural activities within the brain, re-fMRI can be used as an efficient and noninvasive way for investigating neuropathological substrates of many neurological and psychiatric disorders at a whole-brain system level (Admon et al, 2012;Ganella et al, 2017;Li et al, 2017).…”

Mitochondrial dysfunction: A hidden trigger of autism?

“…Mitochondrial dysfunction could be caused by primary dysfunction that is caused by a mutation in a gene which directly participates in ATP production and secondary dysfunction caused due to other genetic, metabolic and biochemical abnormalities that affect the ability of mitochondria to generate the ATP. 21 TWAS (Transcriptome-wide association studies) has been used to recognize the patient risk genes of complex traits 24 and this approach has been applied to autism as well as 25 since biomarkers for autism remain inconsistent. 26 An ongoing meta-examination demonstrated that mitochondrial disease is found in around 5% of children with autism, and 30%–50% of them have biomarkers for mitochondrial abnormality.…”

A Novel Unit-Based Personalized Fingerprint Feature Selection Strategy for Dynamic Functional Connectivity Networks

“…A functional connectivity network (FCN) is usually represented as a graph, where each node represents a brain ROI and the edge between two nodes encodes the associated FC. Nowadays, many FCN models have been developed, from the simple conventional FCN ( Zhang et al, 2015 , 2016 ; Qiao et al, 2018 ) to the complex time-frequency analysis based dynamic FCN ( Yaesoubi et al, 2015 ; Du et al, 2018 ), for diagnosing some neurodevelopmental diseases including the autism spectrum disorder (ASD) ( Fornito et al, 2015 ; Liu et al, 2016 ; Huang et al, 2018 ; Zhao et al, 2018 ), the major depressive disorder ( Greicius et al, 2007 ; Cullen et al, 2014 ) and so on.…”