Those who are interested in translational research of utilizing magnetic resonance imaging techniques in mental disorders.

Obsessive-compulsive disorder (OCD) is a common, heritable and disabling neuropsychiatric disorder [1]. Recent advances in resting-state functional magnetic resonance imaging (rs-fMRI) have facilitated the models of OCD pathophysiology that encompass a specific network of cortico-striato-limbic regions [2]. However, the vast majority of these studies published so far have been based on average differences between groups. Whether functional neuroimaging could be used to inform the clinical assessment of individual OCD patients remains unclear. Thus, the aim of the current study was to apply a novel multivariate pattern analysis technique known as relevance vector regression (RVR) [3-5] to evaluating the potential of rs-fMRI for making accurate predictions about symptom progression in a relatively large sample of drug-naive patients with OCD.The study was approved by the local ethical committee and written informed consent was obtained from all subjects. A total of 68 drug-naive OCD patients and 68 age, sex, handedness and years of education well matched healthy control subjects (HCS) were recruited in current study. The diagnoses of OCD patients were determined by using the structured clinical interview patient edition according to DSM-IV. Clinical symptoms were evaluated using the Yale-Brown Obsessive-Compulsive Scale (Y-BOCS). The MRI examinations were performed via a 3-Telsa GE MRI system with an 8 channel phase array head coil. The rs-fMRI sensitized to changes in the blood oxygen level dependent (BOLD) signal levels were obtained via a GE-EPI sequence (TR/TE=2000/30msec, flip angle=90°, slice thickness=5mm with no gap, 30 axial slices, 200 volumes in each run). Subjects were instructed to relax with their eyes closed without falling asleep during MR examination. The amplitude of low-frequency fluctuation (ALFF) maps were calculated using DPARSF software [6](http://www.restfmri.net). Initially, we performed a standard univariate analysis to explore alterations of regional neural function by comparing ALFF maps between OCD patients and HCS with the voxel-based two-sample t-test in SPM8 (http://www.fil.ion.ucl.ac.uk/spm). Pearson correlation analyses were performed to identify the association between functional neural correlates and OCD symptom severity evaluated by Y-BOCS and subscale scores. In addition, we examined the relationship between symptom scores and the ALFF using multivariate RVR as implemented in PRoNTo (http://www.mlnl.cs.ucl.ac.uk/pronto/) running under Matlab (Mathworks, 2010 release).Relative to HCS, OCD patients showed lower ALFF in the right middle temporal gyrus (MTG) and higher ALFF in the bilateral inferior frontal gyri extending to the anterior insula, bilateral middle frontal gyri, bilateral superior frontal gyri and the left anterior cingulate gyrus (P < 0.05, with family wise error correction) (Figure A). ALFF in right MTG was negatively correlated with compulsive subscale (r = -0.288, P = 0.017) (Figure B). Meanwhile, the application of RVR to rs-fMRI data allowed quantitative prediction of Y-BOCS scores with statistically significant accuracy (correlation = 0.38, P = 0.005; mean squared error = 25.93, P = 0.005) (Figure C). Accurate prediction was based on functional activation in a number of prefrontal, parietal, temporal and occipital regions (Figure D).Our study demonstrated the first evidence that functional neuroimaging techniques might inform the clinical assessment of OCD patients by providing an accurate and objective quantitative estimation of clinical scores. Furthermore, the contribution of posterior brain circuitry (including temporo-parieto-occipital associative areas) provided additional explanation to the well known frontal-subcortical mechanism of OCD.