Selected Publications
Human Brain State Classification via Permutation Entropy of EEG Phase Dynamics across Consciousness Levels and Inattentive-Type ADHD
Athokpam Langlen Chanu, Youngjai Park, Jaesung Choi, Younghwa Cha, UnCheol Lee, Joon-Young Moon, and Jong-Min Park (2026)
Chaos, Solitons & Fractals 209, 118416.
We investigate the complexity of human brain dynamics by applying permutation entropy to dominant EEG phase dynamics, as revealed by the RPA introduced in the previous study. By focusing on a principal anterior–posterior phase mode revealed by RPA rather than raw EEG signals, we obtain a compact measure that converges rapidly and robustly characterizes distinct brain states. The phase-based permutation entropy systematically changes across levels of consciousness and between eyes-open and eyes-closed resting states. These findings demonstrate how the temporal complexity of large-scale phase dynamics can provide a low-dimensional representation of human brain states and complement conventional EEG-based measures.
Sub-Second Fluctuations Between Top-Down and Bottom-Up Modes Distinguish Diverse Human Brain States
Youngjai Park, Younghwa Cha, Hyoungkyu Kim, Yukyung Kim, Jae Hyung Woo, Jehyeop Lee, Hanbyul Cho, George A. Mashour, Ting Xu, UnCheol Lee, Seok-Jun Hong, Christopher J. Honey, and Joon-Young Moon (2026)
Current Biology 36(14), 3548–3565.e8.
We introduce Relative Phase Analysis (RPA), a framework that reveals the real-time directionality of large-scale human brain activity at sub-second resolution, uncovering fast brain dynamics that were previously inaccessible to conventional analyses. Using RPA, we show that the human brain spontaneously alternates between top-down and bottom-up directional modes on a timescale of approximately 200 ms. These rapid fluctuations are most prominent during wakefulness, progressively diminish under general anesthesia, and are altered in ADHD, demonstrating their fundamental relevance across diverse brain states. Simultaneous EEG–fMRI further reveals that top-down dynamics preferentially engage higher-order cognitive networks, whereas bottom-up dynamics are associated with sensory systems. Connectome-based oscillator models reproduce these transitions, suggesting that rapid alternation between top-down and bottom-up modes constitutes a fundamental organizing principle of whole-brain dynamics.
Sustained Attention Task (gradCPT) Dataset Using Simultaneous EEG-fMRI and DTI
Younghwa Cha, Yeji Lee, Eunhee Ji, SoHyun Han, Sunhyun Min, Hyoungkyu Kim, Minseo Cho, Hae Seong Lee, Youngjai Park, and Joon-Young Moon (2026)
Scientific Data 13, 573.
We present a multimodal neuroimaging dataset designed to investigate moment-to-moment fluctuations in human attention. Participants performed the gradual-onset continuous performance task (gradCPT) while simultaneous EEG and fMRI were recorded, together with diffusion-weighted imaging for structural connectivity. The dataset enables neural dynamics to be examined across electrophysiological, hemodynamic, and structural scales within the same individuals. It provides a resource for studying sustained attention, brain-state transitions, multimodal EEG–fMRI relationships, and the links between fast neural dynamics and large-scale brain networks.
Electroencephalogram Correlates of Delayed Emergence After Remimazolam-Induced Anesthesia Compared to Propofol
Yeji Lee, Sujung Park, Hyoungkyu Kim, Youngjai Park, UnCheol Lee, Jeongwook Kwon, Bon-Nyeo Koo, and Joon-Young Moon (2025)
Anesthesia & Analgesia 141(6), 1399–1410.
We investigate the neural dynamics underlying delayed emergence from remimazolam anesthesia by comparing prefrontal EEG activity during remimazolam- and propofol-based general anesthesia. Remimazolam was associated with distinct changes in spectral power and phase-based functional connectivity during recovery, together with slower behavioral recovery. Importantly, phase lag entropy and phase lag index measured during deep anesthesia were associated with subsequent recovery time. These results demonstrate that EEG dynamics can reveal drug-specific trajectories of consciousness recovery and may provide neurophysiological markers for delayed emergence from general anesthesia.
Inter-Regional Delays Fluctuate in the Human Cerebral Cortex
Joon-Young Moon, Kathrin Müsch, Charles E. Schroeder, Taufik A. Valiante, and Christopher J. Honey (2024)
eLife 13: RP92459.
We measure changes in global oscillations and inter-regional couplings from human ECoG recordings. The strength and delays of inter-regional couplings continuously fluctuate in the presence of auditory narrative stimulus. Increases in low-frequency power (4-14Hz) were associated with stronger and more delayed inter-regional couplings, whereas increases in high-frequency power (65+ Hz) were assoicated with weaker couplings and zero-lag delays. Computational models suggest increases in latency could be explained by global increases in the effective influence of inter-regional signaling. These coupling dynamics can reflect a cortex-wide modulation of the relative influence of top-down and bottom-up signals in the human cerebral cortex.
Phase and Amplitude Dynamics of Coupled Oscillator Systems on Complex Networks
Jae Hyung Woo, Christopher J Honey, and Joon-Young Moon (2020)
Chaos 30, 121102.
As a continuation of previous study (Chaos 29), we now investigate phase and amplitude dynamics of the coupled oscillator system. The expansion is made to include amplitude dynamics. By applying mean-field theory, we find in oscillator systems in which individual nodes can independently vary their ampitude over time, qualitatively different dynamics can be produced via vaiations on the coupling form and the underlying network structure. By applying coupled oscillator systems on brain networks, we predict that there are four possible modes of information flow: high/low activity top-down modes and high/low activity bottom-up modes.
Various Synchronous States Due to Coupling Strength Inhomogeneity and Coupling Functions in Systems of Coupled Identical Oscillators
Junhyeok Kim*, Joon-Young Moon*, UnCheol Lee, Seunghwan Kim, and Tae-Wook Ko (2019)
Chaos 29, 011106.
We investigate phase dynamics of the coupled oscillator system on various complex networks. Applying mean-field theory, we find in oscillator systems with coupling strength inhomogeneity qualitatively different dynamics can emerge depending on their underlying network structure. Hub nodes can either phase-lead (becoming source of the information), or phase-lag (becoming sink of the information). This study is the basis for the expansion to include amplitude dynamics and applications to brain networks in Chaos, 30.
Mechanisms of Hysteresis in Human Brain Networks During Transitions of Consciousness and Unconsciousness: Theoretical Principles and Empirical Evidence
Hyoungkyu Kim*, Joon-Young Moon*, George A. Mashour, and UnCheol Lee (2018)
PLoS Computational Biology 14(8), e1006424.
Hysteresis, characterized by distinct forward and reverse phase transitions, is ubiquitous phenomena apearing in many complex systems of the nature. We take a network-based approach and show that anesthetic state transitions share the same underlying mechanism of other hysteresis in nature: nucleation and percolation. Indeed, via computational modeling, analytic study, and human EEG analysis, we show various hysteresis phenomena of conscious-unconscious transition can be explained by genetic network features. This study provide a possbility for a unified framework for radically different conscious state transitions associated with sleep, anesthesia and disorders of consciousness.
Structure Shapes Dynamics and Directionality in Diverse Brain Networks: Mathematical Principles and Empirical Confirmation in Three Species
Joon-Young Moon, Junhyeok Kim, Tae-Wook Ko, Minkyung Kim, Yasser Inturria-Medina, Jee-Hyun Choi, Joseph Lee, George A. Mashour, and UnCheol Lee (2017)
Scientific Reports 7, 46606.
As a continuation of previous study (PloS Comp. Biol. 11), we refine our mathematical principle explaining the emergence of directionality from the underlying brain network structure. We apply our methods to brain networks of human, macaque, and mouse, successfully predicting information flow dynamics of the empirical EEG/ECoG data. The unique global directionality patterns in resting state brain networks of each species can be predicted by their distinctive brain network structures. This comprehensive study forms a foundation for a understanding of how neural information is drected and integrated in complex brain networks across three mammalian species.
General Relationship of Global Topology, Local Dynamics, and Directionality in Large-Scale Brain Networks
Joon-Young Moon, UnCheol Lee, Stefanie Blain-Mraes, and George A. Mashour (2015)
PloS Computational Biology 11(4), e1004225.
How does the brain network organization determine local functions and information transfer patters? In this study, we show inter-node directionality arises naturally from the topology of the network. Analytical, computational, and empirical results all demonstrate that, on average, network nodes with more connections lag in phase, while ower-degree nodes lead. We demonstrate our novel phase analysis method can predict the directionality patterns in human brain networks across different states of consciousness. Our findings provide a straighforward method to dissect how directionality between interaction node is shapes in brain networks. Furthermore, the underlying mathematical relationship between node connections and directionality patterns has the potential to advance network science across multiple disciplines.