The Evolution of Patient-Centered Nursing Care: From Passive Recipients to Active Partners in the Era of Generative AI.

Alzheimer’s disease (AD) is a progressive neurodegenerative condition marked by multi-domain cognitive deterioration spanning memory, executive function, language, and visuospatial abilities [,]. As the leading cause of dementia, AD accounts for approximately 70% of cases worldwide and imposes a growing burden on patients, families, and healthcare systems [,]. Many individuals first transition through a prodromal phase termed mild cognitive impairment (MCI), in which cognitive decline exceeds that expected with normal aging but remains below the clinical threshold for dementia [,]. Despite extensive research, the neurophysiological processes linking stage-specific cognitive deterioration to altered brain network activity remain incompletely understood. Electroencephalography (EEG) non-invasively captures cortical dynamics at the millisecond timescale, a resolution well suited to the neural computations that underlie cognition []. Applying quantitative signal-processing methods to standard EEG recordings, an approach known as quantitative EEG (qEEG), enables neurophysiological biomarkers to be extracted with greater objectivity and precision [,]. Among qEEG techniques, resting-state EEG (rEEG) microstate analysis has attracted particular interest for characterizing the temporal organization of large-scale brain networks in relation to cognitive function [,,]. rEEG microstates are commonly defined as quasi-stable scalp potential topographies, typically derived from alpha-band activity (8–12 Hz), each persisting for approximately 60–120 ms before transitioning to another state [,]. The widely adopted four-class solution delineates Classes A through D, corresponding, respectively, to auditory-phonological, visual, salience, and frontoparietal executive control networks [,,], whereas extended solutions identify up to seven classes, with Classes E, F, and G capturing supplementary configurations []. Notably, Classes D and E exhibit substantial functional overlap and are frequently used interchangeably, with label assignment contingent on the clustering approach and the number of extracted classes [,]. These repeated states of brain electrical activity are thought to capture momentary states of spontaneous mental processing. In this sense, rEEG microstates have been described as “atoms of thought,” as they may represent short-lived but meaningful episodes in the organization of neural information processing []. Consistent with this interpretation, a convergent body of evidence has demonstrated robust associations between microstate parameters and a variety of cognitive and perceptual operations [,], underscoring the theoretical and translational significance of microstate analysis for the study of neurological disorders. The temporal organization of microstate dynamics is commonly characterized through three principal parameters: duration, occurrence, and coverage [,]. Duration refers to the mean time a microstate configuration remains stable before transitioning to a subsequent state and has been proposed to reflect the efficiency and flexibility of large-scale neural network engagement during spontaneous brain activity []. Occurrence denotes the rate at which a given microstate is reinstated per unit time and is understood to index the brain’s propensity to recruit specific resting-state networks [,]. Coverage represents the proportion of total recording time occupied by each microstate class and, by integrating both duration and occurrence, provides a composite measure of each network state’s overall dominance within the resting-state EEG signal [,].