| There are obvious individual differences in the response of patients with major depressive disorder(MDD)to the same treatment they received.Researchers have found brain imaging indicatoors including magnetic resonance imaging(MRI)markers that can be used to predict treatment response in different MDD patients.However,it is not clear whether these indicators change before and after treatment and the relationship between the changes of indicators and the changes of clinical status.In the present study,MRI and pattern classification techniques are proposed to explore this issue,to investigate the stability of the individually predicative MRI markers obtained by scanning before treatment.MRI scans were performed respectively before and after the treatment on thirty-one patients with MDD,and the interval was 12 weeks during which patients were uniformly treated with Selective Serotonin Reuptake Inhibitor(SSRI)antidepressants.According to Jiang's criteria,patients whose scores of HAMD-24 changed more than50% before and after treatment and also whose scores were less than 10 after treatment were classified as remitters(n=16),while those who did not meet this criteria were classified as non-remitters(n=15).The MRI data obtained before treatment were retrospectively analyzed using the DPABI toolkit to find brain imaging indicators which were significantly different between the two groups.Using support vector machine combined with pattern classification,the indicators as features were then extracted and applied to pattern classification to establish a classification model for predicting individual patient's responsiveness to treatment.The paired-sample t-test and intra-group consistency analysis was used for longitudinal analysis of preand post-treatment brain imaging data of each group.Pearson correlation analysis method was used to analyze the relationships between the change of indicators and the change of HAMD-24 score before and after treatment.The results showed that there was no significant difference in structral MRI before treatment.For the rs-f MRI,the ALFF value of the right inferior frontal gyrus(IFG)in the non-remission group were significantly lower than that in the remission group,and the Re Ho value of the left middle temporal gyrus(MTG),the DC value of the left middle occipital gyrus(MOG)in the non-remission group were significantly higher than that in the remission group.The accuracy of classification model was up to87.10%.The result of the paired-samples t-test showed that as features of the pattern classification,there was no significant difference between pre-and post-treatment imaging indicators in each group.Internal consistency analysis found that the ALFF of the right inferior frontal gyrus(IFG)and the Re Ho of the left middle temporal gyrus(MTG)were significantly consistent before and after treatment,while the DC of left middle occipital gyrus(MOG)was not significantly consistent before and after treatment.Pearson correlation analysis found that whether in the remission group or in the non-remission group,the changes of ALFF and the changes of Re Ho were not significantly correlated with the changes of HAMD-24 scores.For DC,the changes of DC were not significantly related to the changes of HAMD-24 scores in the remission group,and the changes of DC were significantly positively related to the changes of HAMD-24 scores in the non-remission group.The results of the present study suggest that among the MRI predictors of treatment response in MDD,the ALFF of the right inferior frontal gyrus(IFG)and the Re Ho of the left middle temporal gyrus(MTG)may be stable trait-dependent markers,whereas the DC of left middle occipital gyrus(MOG)may be unstable state-dependent marker.Findings from this study may aid clinicians in improving diagnosis and tailoring more suitable therapy to each patient. |
PDF Full Text Request