BackgroundIn order to reduce unsuccessful treatment trials for depression, neuroimaging and genetic information can be considered as biomarkers. Together with machine‐learning methods, prediction models have proved to be valuable for baseline prediction.PurposeTo propose an ensemble learning modeling framework that integrates imaging and genetic information for individualized baseline prediction of early‐stage treatment response of antidepressants in major depressive disorder (MDD).Study TypeProspective.SubjectsIn all, 98 inpatients with MDD.Field Strength/Sequence3.0T MRI and gradient‐echo echo‐planar imaging sequence.AssessmentParticipants were divided into responders and nonresponders based on reducing rates of HDRS‐6 after early‐stage treatment of 2 weeks. Fourteen brain regions of interest were selected according to previous studies. An ensemble learning modeling framework was used to integrate imaging data and genetic data.Statistical TestsSupport vector machine (SVM) with linear kernel was utilized to integrate multimode information and then to construct the prediction model. Leave‐one‐out cross‐validation (LOOCV) was used to evaluate the performance. The position characteristics obtained through SVM‐RFE (recursive feature elimination) algorithm and LOOCV was considered to compare each feature's relative importance for the prediction model.ResultsCompared with the single‐level prediction model, the ensemble learning prediction model showed improvement in prediction performance (accuracy from 0.61 to 0.86 with imaging data and genetic data). Integrated with 14 priori brain regions, the region of interest (ROI) map ensemble learning prediction model can achieve a performance that is analogous with the model with information from whole‐brain regions (both with accuracy of 0.81). The integration of genetic features further improved the sensitivity of prediction (sensitivity from 0.78 to 0.87 under the ensemble learning framework).Data ConclusionOur ensemble learning prediction model demonstrated significant advantages in interpretability and information integration. The findings may provide more assistance for clinical treatment selection in MDD at the individual level.Level of Evidence: 1Technical Efficacy: Stage 2J. Magn. Reson. Imaging 2020;52:161–171.

In major depressive disorder (MDD), the anterior cingulate cortex (ACC) is widely related to depression impairment and antidepressant treatment response. The multiplicity of ACC subdivisions calls for a fine-grained investigation of their functional impairment and recovery profiles. We recorded resting state fMRI signals from 59 MDD patients twice before and after 12-week antidepressant treatment, as well as 59 healthy controls (HCs). With functional connectivity (FC) between each ACC voxel and four regions of interests (bilateral dorsolateral prefrontal cortex [DLPFC] and amygdalae), subdivisions with variable impairment were identified based on groups' dissimilarity values between MDD patients before treatment and HC. The ACC was subdivided into three impairment subdivisions named as MedialACC, Dis-talACC, and LateralACC according to their dominant locations. Furthermore, the impairment pattern and the recovery pattern were measured based on group statistical analyses. DistalACC impaired more on its FC with left DLPFC, whereas Lat-eralACC showed more serious impairment on its FC with bilateral amygdalae. After treatment, FCs between DistalACC and left DLPFC, and between LateralACC and right amygdala were normalized while impaired FC between LateralACC and left amygdala kept dysfunctional. Subsequently, FC between DistalACC and left DLPFC might contribute to clinical outcome prediction. Our approach could provide an insight into how the ACC was impaired in depression and partly restored after antidepressant treatment, from the perspective of the interaction between ACC subregions and critical frontal and subcortical regions.

No abstract