Multivariate Pattern Classification of Primary Insomnia Using Three Types of Functional Connectivity Features

Objective: To explore whether or not functional connectivity (FC) could be used as a potential biomarker for classification of primary insomnia (PI) at the individual level by using multivariate pattern analysis (MVPA). Methods: Thirty-eight drug-naive patients with PI, and 44 healthy controls (HC) underwent resting-state functional MR imaging. Voxel-wise functional connectivity strength (FCS), large-scale functional connectivity (large-scale FC) and regional homogeneity (ReHo) were calculated for each participant. We used support vector machine (SVM) with the three types of metrics as features separately to classify patients from healthy controls. Then we evaluated its classification performances. Finally, FC metrics with significant high classification performance were compared between the two groups and were correlated with clinical characteristics, i.e., Insomnia Severity Index (ISI), Pittsburgh Sleep Quality Index (PSQI), Self-rating Anxiety Scale (SAS), Self-rating Depression Scale (SDS) in the patients' group. Results: The best classifier could reach up to an accuracy of 81.5%, with a sensitivity of 84.9%, specificity of 79.1%, and area under the receiver operating characteristic curve (AUC) of 83.0% (all P < 0.001). Right anterior insular cortex (BA48), left precuneus (BA7), and left middle frontal gyrus (BA8) showed high classification weights. In addition, the right anterior insular cortex (BA48) and left middle frontal gyrus (BA8) were the overlapping regions between MVPA and group comparison. Correlation analysis showed that FCS in left middle frontal gyrus and head of right caudate nucleus were correlated with PSQI and SDS, respectively. Conclusion: The current study suggests abnormal FCS in right anterior insular cortex (BA48) and left middle frontal gyrus (BA8) might serve as a potential neuromarkers for PI.