Recherche Of Individualized Diagnosis Of Major Depressive Disorder And Individualized Prediction Of Antidepressant Efficacy Via Analysis Of Thalamic MRI Features

Magnetic resonance imaging(MRI)studies have found thalamic abnormalities in major depressive disorder(MDD).Although there are significant differences in the structure and function of the thalamus between MDD patients and healthy controls(HCs)at the group level,it is not clear whether the structural and functional features of the thalamus are suitable as diagnostic prediction and treatment prediction aids at the individual level.Here,we tested the predictive value of gray matter density(GMD),gray matter volume(GMV),amplitude of low-frequency fluctuations(ALFF),and fractional amplitude of low-frequency fluctuations(f ALFF)in the thalamus using multivariate pattern analysis(MVPA)methods.A total of 118 subjects(74 MDD patients and 44 HCs subjects)were recruited.39 MDD patients and 35 healthy control subjects underwent MRI scanning twice.Between two scans,the MDD group received Selective Serotonin Reuptake Inhibitors(SSRIs)treatment for three months,the HC group didn’t receive any treatment.The Gaussian process classifier(GPC)was trained to separate MDD patients from HCs,and Gaussian process regression(GPR)was trained to predict depression scores and the percentage decrease in HAMD score after treatment.The main findings were as follows:(1)The balanced accuracy of the GPC trained with thalamic GMD was 96.59%(P=0.001).The accuracy of the GPC trained with thalamic GMV was 94.32%(P<0.001).The correlation between HAMD score targets and predictions in the GPR trained with GMD was 0.90(P=0.001,r~2=0.82),and in the GPR trained with GMV,the correlation between HAMD score targets and predictions was 0.89(P<0.001,r~2=0.79).(2)The results after three months of treatment showed a significant correlation between true HAMD score and predictions(P=0.031,r~2=0.05)in GPR trained with GMD,but not in GPR trained with GMV(P=0.16,r~2=0.00).(3)The models trained with ALFF and f ALFF in the thalamus failed to discriminate MDD patients from HC participants,and also failed to predict the efficacy of antidepressants.Findings from this study suggested that GMD and GMV,but not functional indicators of the thalamus,have good potential for the individualized diagnosis of MDD and prediction of efficacy of antidepressant therapy.To our knowledge,this is the first study to focus on the thalamus to predict MDD using machine learning methods at the individual level.