Multimodal Neuroimaging in the Prediction of Deep TMS Response in OCD

<b>Backgrounds:</b> Brain morphological biomarkers could contribute to understanding the treatment response in patients with obsessive-compulsive disorder (OCD). Multimodal neuroimaging addresses this issue by providing more comprehensive information regarding neural processes and structures. <b>Objectives.</b> The present study aims to investigate whether patients responsive to deep Transcranial Magnetic Stimulation (TMS) differ from non-responsive individuals in terms of electrophysiology and brain morphology. Secondly, to test whether multimodal neuroimaging is superior to unimodal neuroimaging in predicting response to deep TMS. <b>Methods.</b> Thirty-two OCD patients who underwent thirty sessions of deep TMS treatment were included in the study. Based on a minimum 50% reduction in Yale-Brown Obsessive Compulsive Scale (Y-BOCS) scores after treatment, patients were grouped as responders (n = 25) and non-responders (n = 7). The baseline resting state qEEG and magnetic resonance imaging (MRI) records of patients were recorded. Independent sample t-test is used to compare the groups. Then, three logistic regression model were calculated for only QEEG markers, only MRI markers, and both QEEG/MRI markers. The predictive values of the three models were compared. <b>Results.</b> OCD patients who responded to deep TMS treatment had increased Alpha-2 power in the left temporal area and increased volume in the left temporal pole, entorhinal area, and parahippocampal gyrus compared to non-responders. The logistic regression model showed better prediction performance when both QEEG and MRI markers were included. <b>Conclusions.</b> This study addresses the gap in the literature regarding new functional and structural neuroimaging markers and highlights the superiority of multimodal neuroimaging to unimodal neuroimaging techniques in predicting treatment response.