The human brain exhibits ongoing large-scale network dynamics even in the absence of external stimuli, reflecting the continuous nature of thought and consciousness. Electroencephalography (EEG) microstates transient, quasi-stable topographical patterns lasting 60–120ms have been proposed as the “atoms of thought,” representing building blocks of spontaneous mental activity [1,2]. Four canonical microstate classes (A, B, C, and D) have been consistently identified across individuals, each linked to distinct functional networks. Microstate A has been associated with phonological and auditory processing, microstate B with visual-sensorimotor networks, microstate C with the default mode network (DMN) and self-referential cognition, and microstate D with attentional control and salience detection [2-4].

Importantly, microstate dynamics are not static but vary according to cognitive states and task demands. For example, increased duration and coverage of microstate C have been observed during internally oriented cognition such as autobiographical memory retrieval and mind-wandering [5,6]. Similarly, microstate A has been shown to dominate during verbal thinking and inner speech, consistent with its association with auditory networks [7]. Alterations in microstate features have also been reported in clinical conditions characterized by disrupted self-referential processing, including depression, schizophrenia, and ADHD [2,8,10]. These findings underscore the sensitivity of microstate parameters to both normative and pathological variations in self-related cognition.

Self-reflection the process of evaluating one’s thoughts, emotions, and experiences provides an ecologically valid way to probe the neural underpinnings of introspection. Neuroscientific studies consistently implicate the DMN in self-reflection, highlighting its role in autobiographical memory, self-awareness, and meta-cognition [10,11]. At the same time, inner speech is often recruited during reflective processes, serving as a cognitive tool for structuring thought and evaluating self-relevant information [12]. Together, these processes suggest that both microstate C (self-referential DMN activity) and microstate a (verbal/phonological processing) may play central roles during periods of deliberate introspection.

Despite this theoretical convergence, few empirical studies have directly investigated EEG microstate dynamics during self-reflection tasks. Most existing work has focused on resting-state or passive mind-wandering, leaving open questions regarding the temporal organization of microstates during active introspective states. Understanding these dynamics may provide new insights into how verbal and self-referential processes interact at the millisecond timescale to support reflective thought.

The present study sought to address this gap by examining EEG microstate parameters following self-reflection prompts. Specifically, participants engaged in 30-second resting periods after answering introspective questions, allowing for the assessment of temporal features (mean duration, occurrence, coverage, explained variance, global field power) and transition probabilities of microstates A–D. Based on prior literature, it was hypothesized that (a) microstate A would show enhanced presence, reflecting heightened verbal and inner speech processes, (b) microstate C would also demonstrate increased duration and coverage, consistent with DMN-driven self-referential cognition, and (c) transition dynamics would reveal coordinated switching between these two microstates, indicating interplay between verbal and introspective networks.

Participants

One female adult participant (age = 22 years) was recruited for this exploratory case study. The participant was healthy, right-handed, and reported no history of neurological or psychiatric illness. Written informed consent was obtained prior to participation, and the study protocol was approved by the Institutional Ethics Committee in accordance with the Declaration of Helsinki [13].

Materials and Self-Reflection Task

To elicit self-reflection, a set of 10 introspective questions was developed based on prior research on self-referential cognition [10,11]. The questions encouraged the participant to engage in reflective evaluation of personal traits, experiences, and values (e.g., “What motivates you most in life?” “What do you consider your biggest strength?”). Each question was presented verbally by the experimenter. Following each response, the participant was instructed to rest quietly with eyes closed for 30 seconds while reflecting on the question.

EEG Data Acquisition

EEG was recorded using a 24 channels medical grade system positioned according to the international 10–20 system. Reference channel was attached at left mastoid. Impedances were kept below 10 kΩ.

Procedure

The experiment consisted of 10 self-reflection trials, one for each introspective question. In each trial:

• The experimenter verbally presented the reflection question.
• The participant responded aloud in a brief statement.
• A 30-second eyes-closed resting-state EEG was recorded immediately after the response, during which the participant was instructed to keep still, minimize blinking, and allow natural reflection.

Data Pre-Processing

EEG data were pre-processed using EEGLAB & MATLAB. Continuous data were visually inspected, and epochs containing muscle artifacts or excessive noise were removed using Automatic Continuous Rejection. Signals were resampled at 256 Hz and band-pass filtered between 1–40 Hz. Independent component analyses (ICA) was used to correct ocular and cardiac artifacts. 2 components were cancelled as they had muscular contamination. Cleaned data were used for microstate analysis.

Microstate Analysis

Microstate segmentation was performed following established procedures [2]. The data were filtered (2–20 Hz) and global field power (GFP) peaks were extracted to identify dominant topographies. A modified k-means clustering algorithm was applied to determine four canonical microstate classes (A, B, C, and D). Individual microstate maps were matched to template maps derived from the participant’s own data.

The following temporal parameters were computed:

• Mean duration (ms): average length of continuous segments of a given microstate.
• Occurrence (Hz): frequency of appearance per second.
• Coverage (%): proportion of total analysis time occupied.
• Explained variance (%): proportion of EEG signal variance accounted for by each microstate.
• Global Field Power (GFP): measure of field strength.

In addition, Markov chain transition probabilities were calculated to evaluate preferred switching patterns between microstate classes.

The microstate analysis revealed four stable topographical configurations (Classes A–D) (See Figure 1). Collectively, these microstates explained 75.41% of the global variance, with individual contributions differing across classes. Microstate A accounted for the highest explained variance (39.37%), followed by microstates B (14.15%), C (12.44%), and D (9.44%). Thus, microstate A emerged as the most dominant configuration in terms of variance explanation.

Temporal parameters across the four classes are presented in Figure 2. Consistent with the explained variance, microstate A exhibited the longest mean duration (~71 ms) and the highest mean occurrence rate (~5.7 appearances/s), as well as the greatest coverage (~39% of the EEG recording). In contrast, microstate D demonstrated the shortest mean duration (~42 ms), the lowest occurrence rate (~4.1 appearances/s), and the smallest coverage (~17%). Microstates B and C occupied intermediate positions for these temporal features, with mean durations of ~47 ms and ~52 ms, respectively.

The mean Global Field Power (GFP), which reflects the strength of neural activity associated with each microstate, was also highest for microstate A (~0.35 μV), followed by microstates B and C (~0.30 μV), while microstate D displayed the lowest GFP (~0.28 μV).

The transition probability analysis further highlighted systematic temporal dynamics between the classes (See Figure 2, bottom right). Transitions from microstate C to A (+14.4) and from B to A (+15.59) occurred more frequently than expected by chance. Conversely, transitions from B to C (−30.51) and from C to D (−24.61) were markedly underrepresented. These findings suggest that microstate A often acts as a central hub in the switching dynamics, whereas transitions directly between microstates B and C or between C and D are less likely.

Figure 1: Spatial topographies of the four identified microstate classes (A–D) with individual explained variance per map.

Figure 2: Temporal parameters of the microstate classes: (a) explained variance (%), (b) mean duration (ms), (c) mean GFP (μV), (d) mean occurrence (appearances/s), (e) coverage (%), and (f) observed minus expected transition probabilities between classes.

The present study examined EEG microstate parameters during short resting-state periods following guided self-reflection, elicited by introspective prompts. Results revealed a clear dominance of microstate A, which exhibited the highest explained variance, the longest mean duration, and the greatest temporal coverage. Microstates C and B demonstrated moderate contributions, while microstate D was least prominent. These findings indicate that self-reflective rest states are characterized by a predominance of verbal/auditory-related neural dynamics with secondary involvement of self-referential and visual networks.

Integration with Prior Literature

EEG microstates are brief, quasi-stable scalp voltage configurations that persist for approximately 80–120 ms before transitioning to another state. They are widely regarded as the electrophysiological correlates of spontaneous mental activity, reflecting rapid alternations between large-scale brain networks [2]. Canonical microstate classes have been consistently mapped onto distinct functional systems:

• Microstate A has been linked to phonological and auditory-verbal processing networks, including superior temporal and language cortices [3,5].
• Microstate B has been associated with visual and sensorimotor integration, particularly within occipital and parietal regions [7].
• Microstate C has shown strong overlap with the default mode network (DMN), including the medial prefrontal cortex and posterior cingulate cortex, and has been related to self-referential processing and introspective mentation [2,20].
• Microstate D has been consistently tied to fronto-parietal attentional control networks, reflecting executive regulation and externally oriented cognition [4].

Temporal features of microstates (e.g., duration, occurrence, coverage) are sensitive to both cognitive states and individual differences [7]. For example, microstate A has been shown to increase during tasks involving verbal rehearsal or inner speech, whereas microstate C predominates during autobiographical memory retrieval, mind-wandering, and reflective thought [5,6,15]. These temporal properties are thought to capture the stability of underlying cognitive operations, with longer durations and greater coverage reflecting sustained neural engagement [16].

Findings in the Context of Self-Reflection

The strong dominance of microstate A in the current study suggests that internal dialogue and verbal recapitulation were the primary modes of neural processing during self-reflection. This finding is consistent with evidence that microstate A duration and coverage increase during verbal thought and meta-cognitive evaluation [6,17]. At the same time, the intermediate contribution of microstate C aligns with the role of the DMN in self-referential and introspective thought [20]. Together, the prominence of A and C supports the idea that self-reflection involves a dynamic interplay between verbal/auditory processing networks and DMN-related self-evaluative processes.

The reduced contribution of microstate D indicates that externally directed attentional control was less engaged, consistent with an inwardly oriented cognitive state. Furthermore, transition analysis demonstrated increased switching between C → A and B → A, suggesting an interdependent relationship between self-referential processing (C), visual/sensorimotor imagery (B), and verbal elaboration (A). This pattern resonates with the global workspace theory of consciousness, which emphasizes the integration of distributed neural states into a coherent stream of thought through rapid alternations [18].

Implications and Neurocognitive Significance

The findings support the interpretation that self-reflection biases the resting-state EEG toward a repertoire dominated by verbal and introspective processes, with microstates A and C functioning as complementary neural modes. This result contributes to the conceptualization of EEG microstates as “atoms of thought” [2], wherein transient but recurring scalp topographies form the temporal basis for higher-order cognition.

Beyond their theoretical implications, the present results highlight the practical utility of microstate analysis for probing the neural correlates of introspection. Previous studies have shown that altered microstate dynamics particularly in classes A and C are associated with psychiatric conditions such as depression and schizophrenia, as well as with contemplative practices such as meditation [19,8,17]. By demonstrating sensitivity to self-reflective states in healthy participants, the current findings strengthen the argument for microstates as potential biomarkers of self-related cognition across both normative and clinical contexts.

In addition, the observed pattern of enhanced A–C transitions highlight how internal dialogue and self-referential evaluation jointly scaffold reflective awareness, supporting dynamic cognition models in which alternating brain states enable the integration of cognitive, perceptual, and affective content [18,19]. Future work may benefit from combining EEG microstate measures with neuroimaging or ecological momentary assessment to more precisely delineate how these transient brain states track the unfolding of self-reflective thought.

In summary, the present study shows that self-reflection shifts resting-state EEG dynamics toward increased microstate a dominance and complementary engagement of microstate C, reflecting a neural architecture that favours verbal elaboration and DMN-driven self-referential processing. These findings provide new evidence for the neurocognitive substrates of introspection and underscore the value of microstate analysis in the study of self-related mental phenomena.

1. Lehmann D, Strik WK, Henggeler B, Koenig T, Koukkou M. Brain electric microstates and momentary conscious mind states as building blocks of spontaneous thinking: I. Visual imagery and abstract thoughts. Int J Psychophysiology. 1998; 29: 1-11.
2. Michel CM, Koenig T. EEG microstates as a tool for studying the temporal dynamics of whole-brain neuronal networks: A review. Neuro Image. 2018; 180: 577-593.
3. Britz J, Van De Ville D, Michel CM. BOLD correlates of EEG topography reveal rapid resting-state network dynamics. Neuro Image. 2010; 52: 1162-1170.
4. Custo A, Van De Ville D, Wells WM, Tomescu MI, Brunet D, Michel CM. Electroencephalographic resting-state networks: Source localization of microstates. Brain Connectivity. 2017; 7: 671-682.
5. Seitzman BA, Abell M, Bartley SC, Erickson MA, Bolbecker AR, Hetrick WP. Cognitive manipulation of brain electric microstates. Neuro Image. 2017; 146: 533-543.
6. Milz P, Faber PL, Lehmann D, Koenig T, Kochi K, Pascual-Marqui RD. The functional significance of EEG microstates: Associations with modalities of thinking. Neuro Image. 2016; 125: 643-656.
7. Khanna A, Pascual-Leone A, Michel CM, Farzan F. Microstates in resting-state EEG: Current status and future directions. Neuroscience and Biobehavioral Reviews. 2015; 49: 105-113.
8. Rieger K, Diaz Hernandez L, Baenninger A, Koenig T. 15 years of microstate research in schizophrenia: Where are we. A meta-analysis. Frontiers in Psychiatry. 2016; 7: 22.
9. Da Cruz JR, Favrod O, Roinishvili M, Chkonia E, Brand A, Mohr C, et al. EEG microstates are a candidate endophenotype for schizophrenia. Nature Communications. 2020; 11: 3089.
10. Northoff G, Heinzel A, De Greck M, Bermpohl F, Dobrowolny H, Panksepp J. Self-referential processing in our brain-A meta-analysis of imaging studies on the self. Neuro Image. 2006; 31: 440-457.
11. Andrews-Hanna JR, Smallwood J, Spreng RN. The default network and self-generated thought: Component processes, dynamic control, and clinical relevance. Annals of the New York Academy of Sciences. 2014; 1316: 29-52.
12. Alderson-Day B, Fernyhough C. Inner speech: Development, Cognitive functions, Phenomenology, and Neurobiology. Psychological Bulletin. 2015; 141; 931-965.
13. World Medical Association. World Medical Association Declaration of Helsinki: Ethical principles for medical research involving human subjects. JAMA. 2013; 310: 2191-2194.
14. Qin Y, He X, Zhang M, Liu Y, Li H, Zhang J. Altered EEG microstates in depression: A resting-state study. Frontiers in Psychiatry. 2020; 11; 102.
15. Diaz BA, Pereda E, Garrido MI, Maestu F. EEG microstates and their non-linear dynamics in healthy subjects. Frontiers in Human Neuroscience. 2021; 15: 56.
16. Zanesco AP, King BG, MacLean KA, Saron CD. Within- and between-person correlates of the temporal dynamics of resting EEG microstates. Neuro Image. 2020; 211: 116631.
17. Zhang Y, Wang J, Chen J, Yang L. EEG microstate alterations during mind-wandering and meditation. Frontiers in Psychology. 2023; 14: 11303265.
18. Mashour GA, Roelfsema P, Changeux JP, Dehaene S. Conscious processing and the global neuronal workspace hypothesis. Neuron. 2020; 105: 776-798.
19. Brechet L, Brunet D, Birot G, Gruetter R, Michel CM, Jorge J. Capturing the spatiotemporal dynamics of self-generated, task-initiated thoughts with EEG and fMRI. Neuro Image. 2019; 194: 82-92.