P325 Disrupted Temporal Dynamics in Stroke: A Criticality Framework for Intrinsic Timescales
Kaichao Wu*1, Leonardo L. Gollo1,2
1Brain Networks and Modelling Laboratory, The Turner Institute for Brain and Mental Health, School of Psychological Sciences, and Monash Biomedical Imaging, Monash University, Victoria 3168, Australia
2Institute for Cross-Disciplinary Physics and Complex Systems, IFISC (UIB-CSIC), Palma de Mallorca, Campus University de les Illes Baleares, Spain.
*Email: kaichao.wu@monash.edu
Introduction
Stroke profoundly disrupts brain function [1-3], yet its impact on temporal dynamics—critical for efficient information processing and recovery—remains poorly understood. Intrinsic neural timescales (INT), which quantify the temporal persistence of neural activity, offer a valuable framework for investigating these dynamic alterations [4,5]. However, the extent to which stroke influences INT and the mechanisms underlying these changes remain unclear.
Methods
This study leverages a longitudinal dataset comprising 15 ischemic stroke patients who underwent resting-state functional MRI at five evenly spaced intervals over six months. INT was computed by estimating the area under the positive autocorrelation function of BOLD signal fluctuations across whole-brain regions [6]. We compared stroke patients' INT values to those of age-matched healthy controls to assess lesion-induced disruptions. Additionally, we analyzed the hierarchical organization of INT across functional networks and examined its relationship with motor recovery, classifying patients into good and poor recovery groups based on clinical assessments. To explore potential mechanisms, we modeled networks of excitable spiking neurons using the Kinouchi & Copelli framework [6,7], investigating the causal relationship between neural excitability and INT within a criticality framework (Fig. 1).
Results
Our findings revealed that stroke patients exhibited significantly prolonged INT compared to healthy controls, a pattern that persisted across all recovery stages. The hierarchical structure of INT, which reflects balanced specialization across brain networks, was markedly disrupted in the early post-stroke phase. By two months post-stroke, differences in INT trajectories emerged between recovery groups, with poor recovery patients displaying abnormally prolonged INT, particularly in the dorsal attention, language, and salience functional networks. These findings align with theoretical predictions from excitable neuron network models, which suggest that stroke lesions may shift the brain’s dynamics toward criticality or even into the supercritical regime (Fig. 1).
Discussion
Our results indicate that stroke-induced INT prolongation reflects increased neural network excitability, pushing the brain toward criticality or even into a supercritical state. The persistent INT abnormalities observed in poorly recovering patients suggest that early-stage INT alterations could serve as prognostic biomarkers for long-term functional outcomes. These findings provide insights into stroke-induced disruptions of brain criticality and highlight the potential of non-invasive neuromodulatory interventions to restore normal INT and facilitate recovery [5]. By advancing our understanding of temporal dynamic changes in stroke, this work sheds light on post-stroke neural reorganization and opens new avenues for targeted rehabilitation strategies using non-invasive brain stimulation.
Figure 1. Figure 1. Stroke lesions prolong intrinsic neural timescales and alter network dynamics, shifting them from a slightly subcritical state (blue) toward criticality (red), with the potential to enter a supercritical state. Near a phase transition, cortical network dynamics can be modeled as a branching process, where intrinsic neural timescales peak at the critical point[6].
Acknowledgements
This work was supported by the Australian Research Council (ARC), Future Fellowship (FT200100942), the Rebecca L. Cooper Foundation (PG2019402), the Ramón y Cajal Fellowship (RYC2022-035106-I) from FSE/Agencia Estatal de Investigación (AEI), Spanish Ministry of Science and Innovation, and the María de Maeztu Program for units of Excellence in R&D, grant CEX2021-001164-M.
References
1.Carrera, E., & Tononi, G. (2014).https://doi.org/10.1093/brain/awu191
2.Park, C.-h., Chang, W. H., Ohn, S. H., et al. (2011). https://doi.org/10.1161/STROKEAHA.110.603846
3.Volz, L. J., Rehme, A. K., Michely, J., et al. (2016). https://doi.org/10.1093/cercor/bhv136
4.Golesorkhi, M., et al. (2021). https://doi.org/10.1038/s41522-021-00447-z
5.Gollo, L. L. (2019). https://doi.org/10.7554/eLife.45089.
6. Wu, K., & Gollo, L. L. (2025).https://doi.org/10.1038/s41522-025-00875-2
7.Kinouchi, O., & Copelli, M. (2006). https://doi.org/10.1038/nphys292
