A predictive study of the efficacy of transcutaneous auricular vagus nerve stimulation in the treatment of major depressive disorder: An fMRI-based machine learning analysis.

BACKGROUND

In order to improve taVNS efficacy, the usage of fMRI to explore the predictive neuroimaging markers would be beneficial for screening the appropriate MDD population before treatment.

METHODS

A total of 86 MDD patients were recruited in this study, and all subjects were conducted with the clinical scales and resting-state functional magnetic resonance imaging (fMRI) scan before and after 8 weeks' taVNS treatment. A two-stage feature selection strategy combining Machine Learning and Statistical was used to screen out the critical brain functional connections (FC) that were significantly associated with efficacy prediction, then the efficacy prediction model was constructed for taVNS treating MDD. Finally, the model was validated by separated the responding and non-responding patients.

RESULTS

This study showed that taVNS produced promising clinical efficacy in the treatment of mild and moderate MDD. Eleven FCs were selected out and were found to be associated with the cortico-striatal-pallidum-thalamic loop, the hippocampus and cerebellum and the HAMD-17 scores. The prediction model was created based on these FCs for the efficacy prediction of taVNS treatment. The R-square of the conducted regression model for predicting HAMD-17 reduction rate is 0.44, and the AUC for classifying the responding and non-responding patients is 0.856.

CONCLUSION

The study demonstrates the validity and feasibility of combining neuroimaging and machine learning techniques to predict the efficacy of taVNS on MDD, and provides an effective solution for personalized and precise treatment for MDD.