AUTHOR=Luo Qian , Liu Weixiang , Jin Lili , Chang Chunqi , Peng Ziwen 

TITLE=Classification of Obsessive-Compulsive Disorder Using Distance Correlation on Resting-State Functional MRI Images

JOURNAL=Frontiers in Neuroinformatics

VOLUME=Volume 15 - 2021

YEAR=2021

URL=https://www.frontiersin.org/journals/neuroinformatics/articles/10.3389/fninf.2021.676491

DOI=10.3389/fninf.2021.676491

ISSN=1662-5196

ABSTRACT=Both Pearson correlation and partial correlation methods have been widely used in resting-state functional magnetic resonance imaging(rs-fMRI) studies. However, they can only measure linear relationship although partial correlation excludes some indirect effects. Recent distance correlation can discover both linear and nonlinear dependencies. Our goal was to use multivariate pattern analysis to compare ability of such three correlation methods to distinguish between patients with obsessive-compulsive disorder (OCD) and healthy control subjects (HCS), so as to find optimal correlation method. The main process includes four steps. Firstly, regions of interest are defined by automated anatomical labeling (AAL). Secondly, functional connectivity (FC) matrices are constructed by three correlation methods. Thirdly, the best discriminative features are selected by support vector machine-recursive feature elimination (SVM-RFE) with a stratified-N-fold cross-validation strategy. Finally, these discriminative features are used to train a classifier. We had a total of 128 subjects, of whom 61 had OCD and 67 were normal. All three correlation methods with SVM have achieved good results, among which distance correlation is the best (Accuracy = 93.01%, Specificity = 89.71%, Sensitivity = 95.08%, AUC = 0.94), followed by Pearson correlation, and partial correlation is the last. The most discriminative brain regions for distance correlation are right dorsolateral superior frontal gyrus, orbital part of left superior frontal gyrus, orbital part of right middle frontal gyrus, right anterior cingulate and paracingulate gyri, left supplementary motor area and right precuneus etc. which are promising OCD biomarkers.