Depression is one of the most prevalent mental disorders globally (Richards, 2011), with a lifetime prevalence of 15 % to 18 % (Bromet et al., 2011). According to a 2023 report by the World Health Organization (WHO), approximately 5 % of the global population is affected by depression (Shin et al., 2016). The disorder is primarily characterized by persistent low mood, loss of interest, and a variety of physical, emotional, and cognitive symptoms (Monroe and Harkness, 2022). Depression significantly impairs daily functioning, diminishes quality of life, and increases the risk of suicide (Mendelson and Tandon, 2016). The WHO ranks depression as the third leading cause of the global disease burden, and it is projected to become the leading cause by 2030 (Michl et al., 2013). Extensive research has explored the pathophysiology of depression, focusing on the monoamine hypothesis (Segal et al., 1974 Oct), inflammation (Miller and Raison, 2015), changes in the hypothalamic-pituitary-adrenal (HPA) axis (Keller et al., 2016), neuroplasticity and neurogenesis (Molendijk et al., 2013), and genetic factors (Flint and Kendler, 2014). In addition, neuroimaging studies have examined the role of abnormal communication within large-scale functional brain networks in individuals with depression (Schmitgen et al., 2019; Kang et al., 2023). Despite significant progress, the pathophysiology of depression remains incomplete.

Transcranial magnetic stimulation combined with electroencephalography (TMS-EEG) is a powerful multimodal imaging technique to investigate cortical responsiveness and regional connectivity with high temporal resolution (Olbrich et al., 2014). Compared to traditional neuroimaging techniques, TMS-EEG uniquely captures the intrinsic properties of specific cortical regions without requiring active engagement from participants (Hill et al., 2016). This method also allows for examining neuronal circuit activity and connectivity across various behavioral and pathological states (Farzan et al., 2016). Numerous studies have comfirmed the ability of TMS-EEG to detect cortical responses to TMS interventions and assess connectivity among cortical regions in both healthy and pathological conditions (Casarotto et al., 2011; Rosanova et al., 2009; Ferrarelli et al., 2008; Rogasch and Fitzgerald, 2012). In a TMS-EEG study, researchers observed changes in event-related spectral perturbations, showing that patients with depression exhibited a reduced intrinsic oscillatory frequency in the pre-motor cortex compared to non-depressive controls (Canali et al., 2015). Voineskos and colleagues analyzed waveforms using global mean field amplitude and found that individuals with depression had heightened cortical reactivity and increased inhibitory responses in the dorsolateral prefrontal cortex (DLPFC) compared to healthy controls (Voineskos et al., 2019). However, most previous studies on depression have focused on analyzing EEG data in the temporal and frequency domains using pre-selected electrodes, which may miss crucial information embedded in the EEG signals and fail to account for the multivariate nature of the data.

Microstate analysis, a widely used EEG representation method, integrates information from all electrodes by characterizing the spatial configuration of electric fields on the scalp, which is visualized as a topographical map of electrical potentials (Poulsen et al., 2018). EEG microstates are recurrent configurations of scalp potentials that remain stable for 60 to 120 milliseconds (da Cruz et al., 2020). This approach allows for the simultaneous analysis of signals from all electrodes, creating a global representation of the brain's functional states (Serrano et al., 2018). As a reliable and reproducible technique, microstate analysis offers high temporal resolution in mapping the organization and dynamics of large-scale cortical oscillations (Michel and Koenig, 2018). A growing body of research has identified anomalous changes in EEG microstates across various affective disorders, including depression, anxiety, and bipolar disorder (Vellante et al., 2020; Al Zoubi et al., 2019; Yan et al., 2021). Taken together, microstate analysis provides a valuable tool for investigating the functionality of brain networks in individuals with depression.

Few studies have employed microstate analysis to explore large-scale brain dynamics based on EEG signals during TMS to date. It is particularly noteworthy that the real-time TMS-EEG design offers a unique advantage in capturing the instantaneous dynamic reorganization of brain microstates. This design enables the synchronous recording of immediate brain network responses at the precise moment of TMS pulse delivery, with millisecond-level temporal resolution (Tremblay et al., 2019; Ilmoniemi and Kičić, 2009). Compared to conventional resting-state EEG microstate analysis, this perturbation-response paradigm allows for the active and causal probing of how specific brain regions respond to stimulation, revealing the real-time dynamic changes and reorganization characteristics of the whole-brain functional network as reflected in microstate sequences (Schaworonkow and Voytek, 2021). This high-temporal-resolution, causality-oriented approach is crucial for uncovering potential pathological mechanisms related to the dynamic stability and plasticity of brain networks in patients with depression. It holds promise for identifying biomarkers that more directly reflect neural circuit dysfunction in the context of depressive disorders. The aim of our study is to provide more reliable and clinically relevant characterizations of EEG microstates associated with depression, while also investigating the correlation between depressive symptoms and these microstates. TMS-evoked EEG responses may offer valuable biomarkers for the identification and understanding of depressive disorders.