Investigating the predictive value of different resting-state functional MRI parameters in obsessive-compulsive disorder

Xuan Bu,Xinyu Hu,Lianqing Zhang,Bin Li,Ming Zhou,Lu Lu,Xiaoxiao Hu,Hailong Li,Yanchun Yang,Wanjie Tang,Qiyong Gong,Xiaoqi Huang
DOI: https://doi.org/10.1038/s41398-018-0362-9
2019-01-01
Translational Psychiatry
Abstract:Previous resting-state functional magnetic resonance imaging (rs-fMRI) studies of obsessive-compulsive disorder (OCD) have facilitated our understanding of OCD pathophysiology based on its intrinsic activity. However, whether the group difference derived from univariate analysis could be useful for informing the diagnosis of individual OCD patients remains unclear. We aimed to apply multivariate pattern analysis of different rs-fMRI parameters to distinguish drug-naive patients with OCD from healthy control subjects (HCS). Fifty-four drug-naive OCD patients and 54 well-matched HCS were recruited. Four different rs-fMRI parameter maps, including the amplitude of low-frequency fluctuations (ALFF), fractional amplitude of low-frequency fluctuations (fALFF), regional homogeneity (ReHo) and functional connectivity strength (FCS), were calculated. Training of a support vector machine (SVM) classifier using rs-fMRI maps produced voxelwise discrimination maps. Overall, the classification accuracies were acceptable for the four rs-fMRI parameters. Excellent performance was achieved when ALFF maps were employed (accuracy, 95.37%, p < 0.01), good performance was achieved by using ReHo maps, weaker performance was achieved by using fALFF maps, and fair performance was achieved by using FCS maps. The brain regions showing the greatest discriminative power included the prefrontal cortex, anterior cingulate cortex, precentral gyrus, and occipital lobes. The application of SVM to rs-fMRI features may provide potential power for OCD classification.
psychiatry
What problem does this paper attempt to address?