Neuronavigation procedure: During the iTBS treatment, the high-resolution T1-weighted scans of each subject will be uploaded to the BrainSight neuronavigation system (version 2.4.11), adjusted, and remodelled into a three-dimensional model. Then, the neuronavigation system is calibrated and validated on the subject. Next, we localized each participant's left-lateral dlPFC stimulation target using Talairach coordinates (-45, 45, 35). The head position is standardized using the nasal tip, sinuses, and bilateral ear screens as baseline markers, and the location of the target on the scalp is identified in each participant's natural space anatomical image. In the end, we observe on the computer screen whether the coil position is placed above the target through the navigation software and artificially control the coil to be tangent to the skull, with the center of the figure-of-eight coil always directly above the target.

Target Localization and Cap Placement Procedure: The stimulation target was the EEG 10-20 coordinate F3, corresponding to the probabilistic location of the left dorsolateral prefrontal cortex (DLPFC). We localized this site using the Beam F3 method and a TMS elastic cap (patent no. 201120150878.5, Wuhan YIRUIDE Medical Equipment Co., Ltd.), and verified it with the F3 electrode on the standard EEG cap. Before the first session, the elastic cap was aligned using anatomical landmarks (nasion, inion, and ears), and the F3 position was marked for consistent coil placement. This standardized approach ensured precise and repeatable targeting across TMS sessions.

iTBS protocol: Intervention with an 80-120% resting threshold was used. In the active group, the left dlPFC was selected as the stimulation site and a stimulation protocol with iTBS parameters was given. This was a neuro-modulation protocol of excitatory sequences, with theta burst stimulation at 50 Hz, repeated at 5 Hz, stimulation for 2 s with an interval of 8 s, for a total of 1,800 pulses, with a treatment time of about 20 min/session, two sessions per day, with a15-30 min rest between sessions, and five consecutive days of treatment per week, with a rest of 2 days, and two weeks of treatment.

Depression severity will be assessed, using 24-item Hamilton Depression Scale (HAMD-24) before and after the iTBS intervention. Changes will be computed as (pre-treatment score - post-treatment score)/pre-treatment score × 100%). We hypothesize that the neuronaviogated group has a better treatment effect indexed by the mean HAMD subtraction rate than the sham-control group.

Neuromodulation-induced iTBS changes will be measured by the probability cueing effects in the additional singleton task with ERP effects. N2pc and Pd components will be used to quantify attentional selection and distractor suppression. These ERP components reflect the neural mechanisms of spatial attention in MDD.