Zotev, V. | Search bioRxiv

bioRxiv2020

Predicting Sex from Resting-State fMRI Across Multiple Independent Acquired Datasets

Sex is an important biological variable often used in analyzing and describing the functional organization of the brain during cognitive and behavioral tasks. Several prior studies have shown that blood-oxygen-level-dependent (BOLD) functional MRI (fMRI) functional connectivity (FC) can be used to differentiate sex among individuals. Herein, we demonstrate that sex can be further classified with high accuracy using the intrinsic BOLD signal fluctuations from resting-state fMRI (rs-fMRI). We adopted the amplitude of low-frequency fluctuation (ALFF), and the fraction of ALFF (fALFF) features from the automated anatomical atlas (AAL) and Powers functional atlas as an input to different machine learning (ML) methods. Using datasets from five independently acquired subject cohorts and with eight fMRI scanning sessions, we comprehensively assessed unbiased performance using nested-cross validation for within-sample and across sample accuracies. The results demonstrated high prediction accuracies for the Human Connectome Project (HCP) dataset (area under cure (AUC) > 0.89). The yielded accuracies suggest that sex difference is embodied and well-pronounced in the low-frequency BOLD signal fluctuation. The performance degrades with the heterogeneity of the cohort and suggests that other factors.e.g. psychiatric disorders and demographics influences the BOLD signal and may interact with the classification of sex. In addition, the results revealed high learning generalizability with the HCP scan, but not across different datasets. The intraclass correlation coefficient (ICC) across HCP scans showed moderate-to-good reliability based on atlas selection (ICC = 0.65 [0.63-0.67] and ICC= 0.78 [0.76-0.80].). We also assessed the effect of scan duration on the predictability of sex and showed that sex differences could be detected even with a short rs-fMRI scan (e.g., 2 minutes). Moreover, we provided statistical maps of the brain regions differentially recruited by or predicting sex using Shapely values and determined an overlap with previous reports of brain response due to sex differences. Altogether, our analysis suggests that sex differences are well-pronounced in rs-fMRI and should be considered seriously in any study design, analysis, or interpretation.

neuroscience↗

bioRxiv2020

Canonical EEG Microstate Dynamic Properties and Their Associations with fMRI Signals at Resting Brain

Electroencephalography microstates (EEG-ms) capture and reflect the spatio-temporal neural dynamics of the brain. A growing literature is employing EEG-ms-based analyses to study various mental illnesses and to evaluate brain mechanisms implicated in cognitive and emotional processing. The spatial and functional interpretation of the EEG-ms is still being investigated. Previous works studied the association of EEG-ms time courses with blood-oxygen-level-dependent (BOLD) functional magnetic resonance imaging (fMRI) signal and suggested an association between EEG-ms and resting-state networks (RSNs). However, the distinctive association between EEG-ms temporal dynamics and brain neuronal activities is still not clear, despite the assumption that EEG-ms are an electrophysiological representation of RSNs activity. Recent works suggest a role for brain spontaneous EEG rhythms in contributing to and modulating canonical EEG-ms topographies and determining their classes (coined A through D) and metrics. This work simultaneously utilized EEG and fMRI to understand the EEG-ms and their properties further. We adopted the canonical EEG-ms analysis to extract three types of regressors for EEG-informed fMRI analyses: EEG-ms direct time courses, temporal activity per microstate, and pairwise temporal transitions among microstates (the latter two coined activity regressors). After convolving EEG-ms regressors with a hemodynamic response function, a generalized linear model whole-brain voxel-wise analysis was conducted to associate EEG-ms regressors with fMRI signals. The direct time course regressors replicated prior findings of the association between the fMRI signal and EEG-ms time courses but to a smaller extent. Notably, EEG-ms activity regressors were mostly anticorrelated with fMRI, including brain regions in the somatomotor, visual, dorsal attention, and ventral attention fMRI networks with no significant overlap for default mode, limbic or frontoparietal networks. A similar pattern emerged in using the transition regressors among microstates but not in self-transitions. The relatively short duration of each EEG-ms and the significant association of EEG-ms activity regressors with fMRI signals suggest that EEG-ms manifests successive transition from one brain functional state to another rather than being associated with specific brain functional state or RSN networks.

neuroscience↗