Repetitive transcranial magnetic stimulation (rTMS) is widely used in the treatment of major depressive disorder (MDD). Stanford Accelerated Intelligent Neuromodulation Therapy (SAINT) is currently the latest effective treatment option that can rapidly antidepressant and alleviate suicidal ideation. However, its mechanism of action is unclear. In this study, we applied regional homogeneity (ReHo) and dynamic functional connectivity (DFC) analyses to investigate the temporal similarity of signal fluctuations and dynamic properties of functional connectivity in 26 MDD patients. ReHo analysis revealed alterations of synchronicity of neuronal oscillations after treatment in the default mode network, subcutaneous nucleus network, frontoparietal network, etc. DFC analysis showed that there were two different states of connectivity. In addition, SAINT also induced increased number of state transitions and enhanced DFC variability in the default mode network, subcutaneous nucleus network, frontoparietal network, etc. Furthermore, correlation analysis revealed a significant negative correlation between DFC variability and baseline scale scores. Finally, we built a machine learning model to predict treatment efficacy based on baseline characteristics and demonstrated that neural activity and brain functional connectivity features at baseline effectively predicted outcomes following SAINT treatment. These findings enhance our understanding of the neurological changes in MDD patients undergoing SAINT, offering potential imaging markers for predicting rTMS treatment efficacy.
© 2025. The Author(s).