Abstract
INTRODUCTION: The human brain exhibits complex functions that emerge from interactions among spatially distributed neural regions. Electroencephalography (EEG) microstate analysis has been widely adopted to capture transient topographies reflecting large-scale network dynamics; moreover, it has been linked to cognitive functions, intrinsic brain networks, and neuropsychiatric disorders. Building on this framework, we recently proposed a novel approach based on instantaneous frequency (IF), defined as the temporal derivative of the instantaneous phase, which characterizes microstates in a dimension distinct from that of conventional amplitude-based microstates by explicitly capturing the phase leading and lagging. Although IF microstates have shown promise in characterizing the pathology and cognitive decline in Alzheimer's, their relevance to normal aging has not been investigated. This study aimed to identify age-group differences in large-scale EEG-dynamic properties using IF microstates. METHODS: We recorded resting-state EEG with eyes closed from 29 younger and 18 middle-aged healthy adults. IF time series were extracted from sensor-level EEG signals in the theta and alpha bands. The IF microstates were identified using a hidden Markov model to ensure temporal continuity in state segmentation. Subsequently, we evaluated the sensor-level spatial distributions, mean dwell times, occupancy, and transition probabilities of the IF microstates and assessed age-group differences using appropriate statistical tests with false discovery rate correction. RESULTS: We identified several IF microstates characterized by frontal IF delay and occipital IF lead, as well as microstates deviating from these patterns. Group comparisons revealed age-group differences in dynamic properties; in the middle-aged group, mean dwell times increased in some states and decreased in others, while occupancy and transition probabilities also exhibited significant changes. DISCUSSION: IF microstate analysis provides a novel and informative perspective on age-group differences in spatiotemporal EEG dynamics. This approach, which is distinct from conventional amplitude-based microstates, may be useful for understanding healthy aging neural mechanisms.