Identifying dynamic functional connectivity biomarkers using GIG‐ICA: Application to schizophrenia, schizoaffective disorder, and psychotic bipolar disorder

Abstract

Functional magnetic resonance imaging (fMRI) studies have shown altered brain dynamic functional connectivity (DFC) in mental disorders. Here we aim to explore DFC across a spectrum of symptomatically-related disorders including bipolar disorder with psychosis (BPP), schizoaffective disorder (SAD) and schizophrenia (SZ). We introduce a group information guided independent component analysis (GIG-ICA) procedure to estimate both group-level and subject-specific connectivity states from DFC. Using resting-state fMRI data of 238 healthy controls (HCs), 140 BPP, 132 SAD and 113 SZ patients, we identified measures differentiating groups from the whole-brain DFC and traditional static functional connectivity (SFC), separately. Results show that DFC provided more informative measures than SFC. Diagnosis-related connectivity states were evident using DFC analysis. For the dominant state consistent across groups, we found 22 instances of hypoconnectivity (with decreasing trends from HC to BPP to SAD to SZ) mainly involving post-central, frontal and cerebellar cortices as well as 34 examples of hyperconnectivity (with increasing trends HC through SZ) primarily involving thalamus and temporal cortices. Hypoconnectivities/hyperconnectivities also showed negative/positive correlations, respectively, with clinical symptom scores. Specifically, hypoconnectivities linking postcentral and frontal gyri were significantly negatively correlated with the PANSS positive/negative scores. For frontal connectivities, BPP resembled HC while SAD and SZ were more similar. Three connectivities involving the left cerebellar crus differentiated SZ from other groups and one connection linking frontal and fusiform cortices showed a SAD-unique change. In summary, our method is promising for assessing DFC and may yield imaging biomarkers for quantifying the dimension of psychosis.

A Multi-Domain Connectome Convolutional Neural Network for Identifying Schizophrenia From EEG Connectivity Patterns

“…Future work could leverage on recent generalizations of CNNs to graphs [52], which can learn representations of graph-structured data lying on irregular or non-Euclidean domains and potentially improve classification of networks. While this study focuses on classifying SZ and HC, further studies could explore application of the MDC-CNN with EEG-based connectome features to automatically discriminate between SZ, schizoaffective disorder and psychotic bipolar disorder which is more difficult in differential diagnosis due to overlapping clinical symptoms [53]. Despite the effectiveness in modeling temporal dependencies in the time-varying functional connectivity, traditional fully-connected LSTM-RNNs suffers the limitation that it takes only vectorized input features which fails to preserve the spatial correlation structure in brain networks.…”

Editorial: Identifying Neuroimaging-Based Markers for Distinguishing Brain Disorders

“…There has been evidence that brain functional connectivity is time-varying, and clustering (e.g., K-means) and decomposition methods can be used to extract connectivity states from dynamic connectivity patterns (Hutchison et al, 2013;Allen et al, 2014;Calhoun et al, 2014;Preti et al, 2017). Most previous dynamic connectivity studies focused on the dynamics of the connectivity between different brain regions or networks Yu et al, 2015;Du et al, 2016Du et al, , 2017Du et al, , 2018a. A study by Bhinge et al proposed a novel approach to measure both the voxelwise spatial variability in functional networks and the dynamic functional network connectivity (dFNC).…”

Detecting the Information of Functional Connectivity Networks in Normal Aging Using Deep Learning From a Big Data Perspective

“…For example, Calhoun et al performed independent component analysis (ICA) to decompose fMRI data into reduced-dimensional time series and spatial patterns, which were further used for modal fusion and classification prediction (Calhoun and Sui, 2016). Du et al (2017) proposed a new scheme of group information-oriented ICAs to mine dynamic functional connectivity from the fMRI data of patients with various mental disorders, detecting the differences of brain function between groups related to diseases, which were not found by traditional static methods. However, traditional linear machine learning algorithms rely on feature extraction, which cannot handle non-linear and complex relationships in data, and lack the ability to process FCs directly as well (Plis et al, 2014).…”