Skip to main content
NCBI home page

This is a preprint.

It has not yet been peer reviewed by a journal.

The National Library of Medicine is running a pilot to include preprints that result from research funded by NIH in PMC and PubMed.

bioRxiv logoLink to bioRxiv
[Preprint]. 2026 Apr 9:2026.04.08.717312. [Version 1] doi: 10.64898/2026.04.08.717312

Mapping developmental patterns of intrinsic timescale

Golia Shafiei 1,2,3, Joëlle Bagautdinova 1,2,3, Valerie J Sydnor 4, Dani S Bassett 5,6,7, Deanna M Barch 8,9, Matthew Cieslak 1,2,3, Yong Fan 10,11, Elizabeth Flook 1,2,3, Alexandre R Franco 12,13,14, Gregory Kiar 12, Audrey C Luo 1,2,3, Michael P Milham 12,13, Linden Parkes 15, Taylor Salo 1,2,3, Leah H Somerville 16, Tien T Tong 1,2,3, Russell T Shinohara 17,18,19, Theodore D Satterthwaite 1,2,3
PMCID: PMC13081997  PMID: 41993353

Abstract

Intrinsic timescale is a commonly used measure of spontaneous neural dynamics that quantifies the temporal window of processing of neuronal populations. Intrinsic timescale displays a hierarchical cortical organization across multiple species and imaging modalities, with shorter timescales in sensorimotor cortex compared to association cortex. However, less is known about how intrinsic timescale evolves during human brain development and whether its cortical maturation patterns generalize to independent developmental samples. Here we estimate the intrinsic timescale in two independent datasets of youth (HCPD: n=565; HBN: n=729; age range 8–22 years) and investigate its neurodevelopmental patterns. We find that developmental changes in the intrinsic timescale follow a hierarchical pattern that recapitulates an axis spanning sensorimotor to association cortices (S–A axis). Our analysis of an independent healthy young adult dataset (HCPYA: n=973, age range 22–37 years) underscores the specificity of these developmental findings, suggesting that the intrinsic timescale develops along the S–A axis in youth and stabilizes in adulthood. Together, these results reveal convergence between major axes of cortical organization and development, highlighting intrinsic timescale as a principled marker of hierarchical brain maturation in youth.

Introduction

Examining the spontaneous neural activity of the human cortex can help reveal how neural populations integrate information over time. Intrinsic timescale is a commonly used measure of spontaneous neural dynamics that captures the temporal window over which neural populations accumulate, maintain, and dissipate information (“memory” of a signal; Hasson et al., 2008; Honey et al., 2012; Murray et al., 2014). Across species, recording techniques, and imaging modalities, intrinsic timescale is hierarchically organized across the cortex in adulthood: sensorimotor regions exhibit short timescales, whereas association cortices show longer integration windows (Murray et al., 2014; Ito et al., 2020; Gao et al., 2020; Raut et al., 2020; Shafiei et al., 2023). This hierarchy aligns closely with major axes of cortical structure, function, and evolution, suggesting shared principles that link local circuit properties to large-scale cortical organization (Margulies et al., 2016; Burt et al., 2018; Huntenburg et al., 2018; Demirtas et al., 2019; Sydnor et al., 2021). Recent developmental findings suggest that the same cortical axes that organize intrinsic timescale in adults, particularly the sensorimotor–association (S–A) axis, also align with brain maturation in youth (Sydnor et al., 2021; Luo et al., 2024; Bero et al., 2026). However, it remains unclear if the development of the intrinsic timescale follows a hierarchical sequence. Here, we evaluate if intrinsic timescale quantified using functional Magnetic Resonance Imaging (fMRI) develops along the cortical hierarchy using multiple large samples of children and adolescents.

Regional variation in intrinsic timescale is thought to arise from underlying cellular and circuit-level mechanisms that shape the temporal integration properties of local neuronal populations (Kiebel et al., 2008; Chaudhuri et al., 2015; Murray et al., 2018; Wang, 2020). At the cellular level, neuronal properties such as intrinsic membrane time constants, ion channel dynamics, and dendritic morphology set a baseline for temporal properties of neuronal populations (Wang, 1999; Chaudhuri et al., 2015; Murray et al., 2018; Wang, 2020). At the circuit level, features such as recurrent connectivity and the balance of excitation and inhibition (E/I balance) strongly shape local dynamics (Chaudhuri et al., 2015; Wang, 1999; 2002; 2008; 2020; Murray et al., 2018). Densely recurrent, feedback-rich circuits can sustain activity over longer periods (yielding longer timescales), whereas predominantly feedforward architectures promote rapid signal decay and shorter timescales. Specifically, strong recurrent excitation, mediated by slow NMDA receptors (Wang, 1999), is thought to maintain the persistent and sustained activity in the absence of immediate sensory stimulation (Wang, 2008; 2020; Chaudhuri et al., 2015; Gao et al., 2020). These properties are not uniformly distributed across the cortex but instead vary systematically along cortical hierarchies, constraining the temporal integration properties of neural populations (Murray et al., 2018; Beul & Hilgetag, 2019). Higher-order association regions generally exhibit longer integration windows than primary sensory areas that have more transient and rapid changes in activity in response to immediate external stimuli (Murray et al., 2014; Shafiei et al., 2020). The sustained activity and longer integration windows observed in association cortex are potentially supported by a combination of local recurrent and feedback circuitry as well as broader system-level interactions that stabilize persistent activity (Wang, 2020; Kauvar et al., 2025).

At the systems level, intrinsic timescales have been found to provide robust and interpretable measures of spontaneous neural dynamics across modalities (e.g. fMRI, ECoG/EEG, single-unit recordings; Watanabe et al., 2019; Gao et al., 2020; Golesorkhi et al., 2021a). Notably, convergent results across modalities support the idea that intrinsic timescale captures a conserved aspect of cortical dynamics. In humans, resting-state fMRI has made intrinsic timescale particularly accessible for mapping these dynamics at the whole-brain level. Previous studies have identified links between measures of intrinsic timescale and cognition and behavior (Watanabe et al., 2019; Gao et al., 2020; Golesorkhi et al., 2021b; Wang, 2022), as well as their implications in brain disorders and diseases (Wengler et al., 2020; Xie et al., 2023). However, less is known about how these dynamics change over development.

Understanding how the intrinsic timescale develops is crucial for several reasons. First, it sheds light on fundamental aspects of developmental neurobiology, particularly the relationship between processing speeds and cognitive performance (Waschke et al., 2021). Acquiring the appropriate processing speed at different developmental stages is essential for the brain’s optimal function, as varying timescales facilitate distinct types of information processing. For example, fast timescales enable efficient transmission of simple information, while slower timescales support the integration of complex, feedback-driven signals (Chaudhuri et al., 2015; Wang, 2008; 2020). This differentiation is vital for cognitive functions such as memory and decision-making (Gao et al., 2020). Furthermore, recent reports indicate that neural dynamics shift across the lifespan (Gao et al., 2020; Truzzi & Cusack, 2023; Wu & Gollo, 2025; Bero et al., 2026), reinforcing the need to explore how these dynamics change with age and how they relate to cognition and disorders. Additionally, observed alterations in intrinsic timescale in developmental disorders, such as autism spectrum disorder, underscore the relevance of characterizing these developmental patterns (Watanabe et al., 2019). By elucidating how intrinsic timescale matures across childhood and adolescence, we can gain insights into both typical and atypical cognitive development.

However, relatively few studies have examined how intrinsic timescale develops across the cortex in youth. Prior developmental work has primarily focused on electrophysiological modalities such as EEG and ECoG (McKeon et al., 2025; Miles et al., 2026), reporting an overall decrease in intrinsic timescale with age. In contrast, fMRI-based investigations in youth remain limited. One study reported a narrowing of intrinsic timescale in the hippocampus, focusing on a specific structure rather than the whole cortex (Zeng et al., 2024). Another study in infants showed that neonates exhibit a markedly different timescale organization than adults, characterized by generally longer timescales and network-specific variability (Truzzi & Cusack, 2023). Notably, this work also demonstrated that the relative ordering of unimodal and transmodal networks differs between infancy and adulthood, suggesting that development involves a redistribution of timescales across cortex rather than a uniform global shift. Despite these observations, the development of fMRI-based intrinsic timescale across the whole cortex in youth remains relatively uncharacterized.

In this study, we map developmental changes in the intrinsic timescale across childhood and adolescence using resting-state fMRI from a large cohort of typically developing youth. We estimate regional intrinsic timescale and characterize its maturation with respect to the cortical hierarchy defined by the S–A axis. We further assess the generalizability of the observed developmental patterns in a large independent developmental sample. We then assess the specificity of our developmental results by also examining associations between intrinsic timescale and age in young adulthood. Together, these analyses reveal that the development of intrinsic timescale aligns with the cortical hierarchy in youth.

Results

Cortical maps of intrinsic timescale were estimated using the autocorrelation function of resting-state fMRI time series for two developmental cohorts (ages 8–22 years) and a young-adult dataset (ages 22–37 years) (Figure 1). We then used generalized additive models (GAMs) to examine linear and nonlinear associations between fMRI intrinsic timescale and age. We also assessed whether age-related changes in intrinsic timescales align with the cortical hierarchy defined by the S–A axis.

Figure 1. Quantifying intrinsic timescale using autocorrelation function |.

Figure 1.

Open in a new tab

Fully processed parcellated fMRI time-series were used to estimate intrinsic timescale for each brain region (i.e., parcel) and participant. We estimated an autocorrelation function (ACF) for the normalized (i.e., z-scored) time series. Intrinsic timescale was then quantified as the sum of positive autocorrelation values (ac > 0; colored parts of the ACFs above) multiplied by TR for each cortical region and individual. This is approximately equivalent to calculating the area under the curve of the positive part of the ACF curve (i.e., shaded areas). As demonstrated in the example plots, a fast-decaying ACF will have shorter memory, and thus a shorter timescale (orange region and ACF plot), compared to a slow decaying ACF (purple region and ACF plot).

Intrinsic timescale is hierarchically organized in youth

We first sought to characterize intrinsic timescale patterns in the two developmental cohorts. Intrinsic timescale varied across the cortex in both developmental datasets: the Human Connectome Project–Development (HCPD; n = 565; Somerville et al., 2018) and the Healthy Brain Network (HBN; n = 729; Alexander et al., 2017) (Figure 2). Visual inspection suggested that the cortical distribution of fMRI intrinsic timescale was broadly segregated into lower-order unimodal regions and higher-order transmodal areas (Figure 2a). The mean timescale maps, averaged across individuals, were positively correlated between the two developmental datasets, demonstrating consistent patterns (Figure 2a; rs = 0.76, pspin = 0.0002). To assess whether timescale maps in youth reflect the cortical hierarchy, we directly compared the average timescale to the S–A axis (Sydnor et al., 2021; Figure 2b). The cortical distribution of timescale was positively associated with the S–A axis in both datasets. However, this association was statistically significant only in HCPD (Figure 2b scatter plots: HCPD: rs = 0.37, pspin = 0.03; HBN: rs = 0.27, pspin = 0.12). Further analysis demonstrated that intrinsic timescale was generally significantly longer in association cortex compared to sensorimotor cortex in both datasets (Figure 2b boxplots: HCPD: t = 7.8, pspin = 0.01; HBN: t = 6.6, pspin = 0.04; two-tailed). Collectively, these findings indicate that the intrinsic timescale is hierarchically organized in youth, with longer timescales observed in association cortex compared to sensorimotor cortex.

Figure 2. Intrinsic timescale in developmental cohorts is hierarchically organized |.

Figure 2.

Open in a new tab

(a) Cortical distribution of intrinsic timescale reflects similar patterns in two independent developmental cohorts: the HCPD and HBN datasets. Intrinsic timescale maps were compared directly between the two datasets (scatter plot). Each brain region (i.e., each circle in the scatter plot) is colored based on the region’s rank along an organizational axis of the cortex that spans sensorimotor to association cortices (S–A axis). (b) Intrinsic timescale in youth (HCPD and HBN dataset) reflects the hierarchical organization captured by the S–A axis, such that sensorimotor regions have shorter timescales while association regions have longer timescales. Significance of the associations between cortical maps (i.e., scatter plots) was assessed using spatial autocorrelation-preserving permutation tests (i.e., spin test). rs denotes Spearman’s rank correlation coefficient. Linear regression lines are added for visualization purposes only. Asterisks in the boxplot denote significant differences in the means (two-tailed t-test).

Neurodevelopmental changes in intrinsic timescale align with the cortical hierarchy

To characterize the patterning of developmental change in intrinsic timescale throughout youth, we examined the relationship between participant age and both whole-brain and regional timescale in HCPD, while controlling for participant sex and in-scanner motion. We found that the whole-brain average timescale increases during development in youth (Figure 3a; partial R2 = 0.047, panova = 1.25 × 10−7). Furthermore, we found that regional neurodevelopmental patterns were heterogeneous across the cortex, with smaller age effects in the sensorimotor cortex and larger age effects in association cortex (Figure 3b). Notably, the effect size of the developmental associations aligned with the cortical hierarchy defined by the S–A axis (Figure 3c; rs = 0.41, pspin = 0.001). Further inspection of regional developmental model fits revealed that timescale increased in association cortex but was stable in sensorimotor regions (Figure 3d). Overall, these findings demonstrate that fMRI intrinsic timescale increases with age in youth, mainly driven by increases in timescale in association cortex. Timescale development followed a hierarchical pattern, recapitulating the S–A axis of cortical organization.

Figure 3. Developmental patterns of intrinsic timescale |.

Figure 3.

Open in a new tab

Generalized Additive Models (GAMs) were used to assess linear and nonlinear age effects in intrinsic timescale. (a) A whole-brain GAM was applied to model age-related changes in cortex-wide average timescale. GAM results demonstrated that average timescale increases during development in youth (partial R2 = 0.047, panova = 1.25 × 10−7). (b) Region-wise GAMs were used to examine developmental changes in intrinsic timescale at the regional level. Age effects quantified as partial R2 are depicted across the cortex, displaying a heterogeneous spatial distribution. (c) Age effects on intrinsic timescale were compared to the S–A axis, identifying a hierarchical pattern of developmental changes in intrinsic timescale along the S–A axis. Significance of the association between age effects and S–A axis rank was assessed using 10,000 spin tests. rs denotes Spearman’s rank correlation coefficient. Linear regression line is added for visualization purposes only. (d) Regional trajectories of developmental changes in intrinsic timescale were obtained from region-wise GAM results (as shown in panel b). Each line corresponds to the model fit of each cortical region and is colored based on the region’s rank along the S–A axis. The results demonstrate that intrinsic timescale in association regions increases during development in youth while it remains relatively stable in sensorimotor regions.

Findings replicate in an independent developmental cohort

We next examined the extent to which these developmental findings were generalizable by replicating our analyses in the HBN dataset. Consistent with HCPD, we found that the average intrinsic timescale increased during development in HBN (Figure 4a; partial R2 = 0.037, panova = 1.13 × 10−7). Regional analysis similarly identified heterogeneous developmental effects across the cortex (Figure 4b). As in HCPD, these developmental effects were aligned with the S–A axis (Figure 4c; rs = 0.19, pspin = 0.02). Developmental fits showed that overall, timescale remained stable in sensorimotor regions but increased in association cortex (Figure 4d). To directly evaluate the degree to which results converged across the two datasets, we compared developmental patterns of intrinsic timescale between HBN and HCPD (Figure 5). We found that maps of associations with age (i.e., partial R2 maps) were similar for the two datasets (Figure 5; rs = 0.30, pspin = 0.0002). Overall, 46.5% of the regions in HCPD and 22.5% of the regions in HBN demonstrated significant age effects after FDR correction. The developmental patterns in these regions had a significant spatial overlap between HCPD and HBN (Figure 5; Dice score = 0.42, pspin = 0.0008; see Figure S1 for age of maximal developmental change in significant regions).

Figure 4. Replication in an independent dataset |.

Figure 4.

Open in a new tab

Developmental patterns of intrinsic timescale were replicated in an independent dataset (i.e., HBN). (a) Consistent with the finding in the HCPD dataset, GAM results demonstrated that the average timescale increases during development in youth in the HBN dataset (partial R2 = 0.037, panova = 1.13 × 10−7). (b) Region-wise GAMs identified heterogeneous age effects (i.e., partial R2) across the cortex. (c) Age effects on intrinsic timescale were hierarchically organized along the S–A axis. Significance of the association between age effects and S–A axis rank was assessed using 10,000 spin tests. rs denotes Spearman’s rank correlation coefficient. Linear regression line is added for visualization purposes only. (d) Regional trajectories of developmental changes in intrinsic timescale were obtained from region-wise GAM results (as shown in panel b). Each line corresponds to the model fit of each cortical region and is colored based on the region’s rank along the S–A axis. Similar to HCPD, the results demonstrate that intrinsic timescale in association regions increases during development in youth while it remains relatively stable in sensorimotor regions.

Figure 5. Consistent age effects across two developmental cohorts |.

Figure 5.

Open in a new tab

Age effects (i.e., partial R2) obtained from the HCPD and HBN dataset were directly compared (scatter plot). Significance of the association between age effects from the two datasets was assessed using 10,000 spatial autocorrelation-preserving permutation tests (i.e., spin test). rs denotes Spearman’s rank correlation coefficient. Linear regression line is added for visualization purposes only. Additionally, the spatial overlap between regions with significant age effects after FDR correction was quantified using Dice score. To assess the statistical significance of the between-dataset spatial overlap, spin tests were used to generate a null distribution of Dice scores (10,000 repetitions). The empirical Dice score was then compared to the null distribution of scores to calculate a p-value for the spatial overlap.

Timescale remains relatively stable in adulthood

To assess whether the observed age-related changes in timescale were specific to childhood and adolescence or extended into young adulthood, we repeated all analyses in an independent sample of young adults (i.e., 22–37 years old) from the Human Connectome Project–Young Adults (HCPYA; van Essen et al., 2013; n = 973) (Figure 6). As expected, intrinsic timescale in young adults reflected the hierarchical organization captured by the S–A axis (Figure 6a; rs = 0.4, pspin = 0.01), such that sensorimotor regions had shorter timescales while association regions had longer timescales (t = 7.5, pspin = 0.02, two-tailed). However, unlike the developmental findings, age analysis demonstrated that the whole-brain average timescale remains relatively stable in adulthood (Figure 6b; partial R2 = −0.004, panova = 0.04). Regional age effects estimated using GAMs were not statistically significant for the majority of brain regions (11.7% of regions with significant age effects; Figure 6c) and did not exhibit significant variability along the S–A axis (Figure 6d; rs = −0.1, pspin = 0.58). Overall, these results emphasize the specificity of the developmental findings in HCPD and HBN to the developmental period of 8 to 22 years old, suggesting that fMRI intrinsic timescale develops along the S–A axis in youth and stabilizes in adulthood.

Figure 6. Stable intrinsic timescale in young adults |.

Figure 6.

Open in a new tab

(a) Intrinsic timescale in young adults (HCPYA dataset) reflects the hierarchical organization captured by the S–A axis, such that sensorimotor regions have shorter timescales while association regions have longer timescales. Asterisk in the boxplot denotes significant difference in the means (two-tailed t-test). (b) However, unlike the developmental cohorts, GAM results demonstrated that the whole-brain average timescale remains relatively stable in adulthood (partial R2 = −0.004, panova = 0.04). (c) Region-wise GAMs identified heterogeneous age effects (i.e., partial R2) across the cortex; however, those age effects were not statistically significant for the majority of brain regions. (d) Unlike the developmental cohorts, age effects on the intrinsic timescale did not reflect the hierarchical organization along the S–A axis during adulthood. Significance of the associations between cortical maps (i.e., scatter plots) was assessed using 10,000 spin tests. rs denotes Spearman’s rank correlation coefficient. Linear regression lines are added for visualization purposes only.

Sensitivity analysis

To evaluate whether our findings were influenced by confounding factors and specific analytical choices, we performed a series of sensitivity analyses. We first examined whether the findings were independent from confounding factors such as brain volume and parcel size. Given that brain size changes during development, total brain volume might impact the average whole-brain intrinsic timescale. However, we only found a weak association between estimated Total Intracranial Volume (eTIV) and intrinsic timescale (Figure S2a). To ensure that the developmental patterns of intrinsic timescale were independent from this weak correlation, we repeated the age analysis while including eTIV as a model covariate. The results were consistent with the original findings (Figure S2b). Additionally, given that parcel-wise time series are calculated as the average time series across vertices within a given parcel, parcel size might directly impact smoothness of parcel time series and their autocorrelation functions. To test this, we compared parcel size and intrinsic timescale for each parcel and each participant in HCPD and found no associations (mean rs = 0.01). Moreover, given that HBN is a more heterogeneous sample compared to HCPD and includes help-seeking individuals with higher levels of psychopathology, we repeated all analyses using two subsets of HBN individuals (each with N=600; this approximately matches the HCPD sample size) with low motion or low psychopathology (i.e., p-factor). The findings were consistent with the full HBN sample (Figure S3). Furthermore, to assess the extent to which the results were influenced by the choice of atlas and number of parcels, we repeated the analysis with a lower resolution Schaefer atlas with 200 parcels in HCPD. The results were consistent with the original findings with the Schaefer 400 atlas (Figure S4). Additionally, to test whether intrinsic timescale estimates were sensitive to methodological choices, we quantified the intrinsic timescale in HCPD using two other approaches: (1) fitting an exponential function to ACF and using the exponential decay constant to estimate intrinsic timescale (Ito et al., 2020); (2) directly using the first zero-crossing point of ACF as intrinsic timescale. Both approaches generated intrinsic timescale maps and developmental patterns consistent with the original method (Figure S5). Finally, to ensure that findings were independent from signal-to-noise ratio (SNR) of the fMRI signal, we estimated temporal SNR (tSNR) as the ratio of the time-series mean to standard deviation for each region and participant in HCPD. Neither the intrinsic timescale nor the age effects (partial R2) were significantly associated with tSNR (Figure S6).

Discussion

We systematically characterized how intrinsic timescale measured using fMRI develops during youth. We found that intrinsic timescales are organized along a hierarchical cortical axis (S–A), with longer timescales in association cortex and shorter timescales in sensorimotor cortex. Furthermore, modeling the regional developmental trajectories of timescale demonstrated that intrinsic timescale development aligned with the cortical hierarchy. Importantly, these developmental patterns were robust, appearing in two large independent developmental cohorts. Moreover, findings in a young adult sample demonstrated that age-related changes in intrinsic timescale stabilizes by early adulthood. Together, these findings position intrinsic timescale as an fMRI measure sensitive to neurodevelopment in youth that helps bridge levels of analysis, linking local circuit dynamics to the brain’s broader hierarchical maturational program.

The observed S–A axis pattern of timescale development aligns with well-established principles of cortical organization. In adults, neuronal timescales are known to increase in length from primary sensorimotor regions to higher-order association cortices, mirroring fundamental anatomical and functional gradients (Murray et al., 2014; Demirtas et al., 2019; Gao et al., 2020; Ito et al., 2020). Recent developmental studies have suggested that similar large-scale cortical axes guide maturation (Sydnor et al., 2021; Luo et al., 2024; Nishio et al., 2024; Bero et al., 2026). Our findings confirm that age-related changes in the fMRI intrinsic timescale recapitulate an organizational hierarchy. As youth develop, their brain dynamics conform to the same sensorimotor-to-association gradient that defines the adult brain architecture. Our results also mirror other metrics of brain functional development. Notably, previous work has shown that the maturation of both spontaneous activity amplitude (i.e., ALFF) (Sydnor et al., 2023) and functional connectivity (Luo et al., 2024) unfold along the S–A axis, indicating a hierarchical refinement of neural activity during youth.

While this work was being prepared, a very recent study reported that age-related changes in temporal scales of resting-state cortical activity reflect maturation of cortical myelination during the human lifespan and align with the S–A axis (Bero et al., 2026). In this study, the temporal scales were quantified using lag-1 temporal autocorrelation (lag-1 TA) of fMRI time series using data from HCPD. Bero et al. (2026) also demonstrated that age-related changes in temporal scales vary across the cortex, such that the temporal scales decrease the most in sensorimotor regions (e.g., visual and somatomotor cortices) while they increase in prefrontal areas (Bero et al., 2026). Our findings provide convergent evidence for how neural dynamics develop in youth. Although our measure of intrinsic timescale conceptually overlaps with the temporal scales defined by Bero et al. (2026), the two measures characterize neural dynamics in different ways. Specifically, the measure of timescale we used takes into account the full shape of the autocorrelation function of fMRI time series, whereas the measure of temporal scales in Bero et al. (lag-1 TA) mainly focuses on the autocorrelation function at short lags (i.e., lag 1). Despite these differences in how fMRI dynamics were defined, both studies converge in showing that the development of neural dynamics aligns with the S–A axis. Additionally, the regional heterogeneity of age-related changes observed across both measures suggests that they capture partially overlapping but distinct aspects of neural dynamics. For example, we found that intrinsic timescales remain stable in sensorimotor regions while they increase in association cortex, whereas Bero et al. found that temporal scales decrease in sensorimotor regions while they increase in prefrontal regions

These subtle but important differences underscore the need to integrate multiple approaches to quantifying neural dynamics across imaging modalities and datasets to obtain a more comprehensive view of brain development in youth. For example, prior research in infants suggests that neonates exhibit a distinct timescale organization compared to adults, characterized by generally longer timescales and network-specific variability (Truzzi & Cusack, 2023). Although this study used the same measure of intrinsic timescale as ours, the age range differed substantially (infancy vs. childhood and adolescence). While longer timescales in neonates may relate to slower baseline information processing, they may also reflect differences in how information is integrated in early development. Notably, this work shows that the relative ordering between unimodal and transmodal cortices differs between infancy and adulthood (i.e., higher timescales in unimodal regions in infancy vs. transmodal regions in adulthood). Reconciling these findings with our results therefore likely requires developmental shifts in the relative ordering of timescales across cortical systems.

Furthermore, prior studies using imaging modalities with higher temporal resolution, such as EEG and ECoG, have reported overall decreases in intrinsic timescale during development (McKeon et al., 2025; Miles et al., 2026). Although similar measures of timescale were used across these studies and our study in fMRI, the overall age-related trends timescale varied across modalities. These differences may reflect the fact that different imaging modalities probe distinct aspects of neural dynamics, as they are sensitive to different frequency ranges and may preferentially capture activity arising from different laminar sources (Scheeringa et al., 2016; Sadaghiani & Wirsich, 2020). Finally, discrepancies across studies, even within the same modality, may also arise from differences in data acquisition, preprocessing pipelines, and analytic approaches. Such methodological variation can influence estimates of resting-state dynamics (e.g., time series properties, cross-correlations, and intrinsic timescale; Li et al., 2024) and should be considered when interpreting findings.

We found that the intrinsic timescale increases with age in youth, particularly in the association cortex. This suggests that neural dynamics in association cortices become more integrative as the brain matures. While faster dynamics may facilitate immediate responses, longer timescales in association cortex are likely critical for supporting higher-order cognitive functions such as working memory (Gao et al., 2020). Convergence with established cortical gradients further indicates that intrinsic timescale follows a common developmental program aligned with the brain’s hierarchical organization. By charting normative timescale development from mid-childhood through adolescence, our study helps contextualize findings from both earlier and atypical development (Truzzi & Cusack, 2023; Watanabe et al., 2019). Specifically, our results suggest that from middle childhood onward, intrinsic timescales adopt an adult-like hierarchical ordering, helping bridge the gap between neonatal and mature states. This normative trajectory is also relevant for understanding neurodevelopmental disorders. For example, autism spectrum disorder has been associated with atypical timescale development, where the typical variation in timescales is disrupted (Watanabe et al., 2019). Establishing how timescales mature in typically developing youth therefore provides an important baseline against which such deviations can be interpreted.

The present findings must be interpreted in the context of several methodological considerations. First, our analysis is cross-sectional, which limits inferences about within-person developmental trajectories. Longitudinal studies will be needed to confirm how intrinsic timescales evolve in the same individuals over time and to assess whether individual differences in timescale trajectories predict neurocognitive outcomes or vulnerability to mental health conditions. Second, our analysis includes childhood, adolescence, and early adulthood (age range of 8–22 years), missing critical information about the developmental patterns of intrinsic timescale in infancy and early childhood. Future studies should seek to combine data collected with similar acquisition parameters across a broader range of ages in development to evaluate this possibility directly. Third, the intrinsic timescale was measured via fMRI signals with a moderate sampling rate (TR=0.8s), which constrains the temporal resolution. Faster neural events or fine-grained timescale differences might be missed due to the coarse temporal sampling of fMRI. Future multimodal studies are required to address the cross-modal correspondence of developmental changes in neural dynamics. Finally, we note that our focus was on cortical timescales at the regional level. However, neural dynamics and their characteristics might vary systematically between different brain structures (Zeng et al., 2024). Developmental changes in fMRI intrinsic timescale remain to be examined in subcortical structures and using finer spatial gradients.

Taken together, our findings suggest that development of intrinsic timescale aligns with the cortical hierarchy, with an increase in timescale in the association cortex that likely reflects an expansion of integrative capacity. These results open new avenues for investigating the mechanistic underpinnings of timescale development. Moving forward, it will be valuable to link these findings to multimodal data and evaluate timescale in the context of developmental processes like synaptic pruning, myelination, or shifts in excitation-inhibition balance. Such work could help determine why association cortices exhibit prolonged integration windows during youth while sensorimotor areas remain stable, providing biological insight into the hierarchical development we observed.

Methods

Participants

We used resting-state functional magnetic resonance imaging (fMRI) data from 3 independent cohorts: (1) a developmental dataset used as a discovery sample; (2) an independent developmental dataset used to replicate the findings; (3) a young adult dataset to evaluate the specificity of our results.

Discovery sample – HCPD:

The discovery developmental sample included resting-state fMRI data from the Human Connectome Project–Development (HCPD; Somerville et al., 2018; n=565, age range 8–22 years, 302 female and 263 male). HCPD is a sample of typically developing children and adolescents recruited in the United States of America across four sites: University of Minnesota, Harvard University, Washington University in St. Louis, and University of California-Los Angeles. All study procedures were approved by a central Institutional Review Board at Washington University in St. Louis.

Imaging data were obtained on 3 Tesla Siemens Prisma platforms at the four sites. All HCPD participants had T1 scans passing quality control (QC). The fMRI data included 26 minutes of resting-state scans per participant in total acquired in four runs with the following acquisition parameters: multiband acquisition, 2mm3, TR=800ms, TE=37ms, flip angle=52°, number of volumes=478. We only included fMRI scans with low in-scanner motion (i.e., mean framewise displacement (FD) < 0.2mm; see “Neuromaging data processing” for details); these QC criteria excluded n=48 individuals from the initial sample of n=635 individuals with resting-state data. Additionally, from the remaining sample of n=587 individuals, we excluded n=20 individuals with medical conditions resulting in gross neurological abnormalities or affecting brain function (e.g., individuals with cancer or leukemia, lead poisoning, sickle cell anemia, accidental poisoning, multiple sclerosis, seizure, and brain injury). Finally, we limited our sample to the age range of 8–22 years old, resulting in a final sample with n=565 individuals.

Replication sample – HBN:

The replication developmental sample included resting-state fMRI data from the Healthy Brain Network (HBN; Alexander et al., 2017; n=729, age range 8–22 years, 279 female, 418 male, 32 other). HBN is a sample of children and adolescents residing in the New York City area (United States of America) that is aimed to represent a diverse sample of healthy and help-seeking individuals. The HBN study was approved by the Chesapeake Institutional Review Board.

Our HBN sample included data acquired at four different acquisition sites: Staten Island (SI - Mobile Scanner), Rutgers University (RU), The City University of New York (CUNY), and Citigroup Biomedical Imaging Center (CBIC). Note that we excluded HBN data from the Staten Island (SI) site from our analysis given that SI’s fMRI data was acquired on a different scanner (1.5T Mobile Scanner) with different acquisition parameters. Image acquisition parameters for functional images collected at the RU, CUNY, and CBIC sites were as follows: multiband acquisition, 2.4mm3, TR=800ms, TE=30ms, flip angle=31°, number of volumes=375 (median; over 97% of scans had 375 volumes, with a small subset containing a different number of volumes). We excluded n=93 participants who did not pass T1 QC from the initial sample of n=1743 with resting-state data (for details, see Shafiei et al., 2025). As for HCPD, we also excluded n=831 participants from the remaining sample of n=1650 due to in-scanner motion (i.e., mean FD < 0.2mm). Finally, same as in HCPD, we limited our sample to the age range of 8–22 years old, resulting in a final sample with n=729 individuals.

Young adult sample to evaluate specificity – HCPYA:

Resting-state fMRI data from healthy young adults were obtained from the Human Connectome Project–Young Adults (HCPYA; van Essen et al., 2013; n=973, age range 22–37 years, 523 female, 450 male). All HCPYA imaging data were collected on a customized 3 Tesla scanner at Washington University (WashU). The study procedures were conducted under Institutional Review Board–approved protocols of the WU–Minn Human Connectome Project consortium (Van Essen et al., 2013).

As for HCPD, all HCPYA participants had T1 scans passing quality control (QC). The fMRI Data included 1 hour of resting-state scans acquired over 2 days (2 scans along opposing R/L and L/R phase encoding directions per day; each scan was approximately 15 minutes long). All functional images were acquired with the following acquisition parameters: multiband acquisition, 2mm3, TR=720ms, TE=33.1ms, flip angle=52°, number of volumes=1200 (median; over 99% of scans had 1200 volumes). As for the developmental data, we excluded n=123 participants from the initial sample of n=1097 individuals (with resting-state data) due to in-scanner motion (i.e., mean FD < 0.2mm), resulting in a final sample with n=973 individuals.

Neuroimaging data processing

All three datasets were preprocessed and post-processed with similar image processing pipelines. The detailed descriptions are provided based on reports automatically generated by the implemented processing tools, and are released under the CC0 1.0 Universal license.

Developmental sample preprocessing (HCPD and HBN):

The two developmental datasets (i.e., HCPD and HBN) were preprocessed with fMRIPrep version 22.0.2 (Esteban et al., 2018), which is based on Nipype 1.8.5 (Gorgolewski et al., 2011).

Anatomical data:

The T1-weighted (T1w) image was corrected for intensity non-uniformity (INU) with N4BiasFieldCorrection (Tustison et al., 2010), distributed with ANTs 2.3.3 (Avants et al., 2008), and used as T1w-reference throughout the workflow. The T1w-reference was then skull-stripped with a Nipype implementation of the antsBrainExtraction.sh workflow (from ANTs). Brain tissue segmentation of cerebrospinal fluid (CSF), white-matter (WM) and gray-matter (GM) was performed on the brain-extracted T1w using fast (FSL 6.0.5.1; Zhang et al., 2001). Brain surfaces were reconstructed using recon-all (FreeSurfer 7.2.0; Dale et al., 1999). The brain mask estimated previously was refined to reconcile ANTs-derived and FreeSurfer-derived segmentations. Volume-based spatial normalization to two standard spaces (MNI152NLin6Asym, MNI152NLin2009cAsym) was performed through nonlinear registration with antsRegistration (ANTs 2.3.3), using brain-extracted versions of both T1w reference and the T1w template.

Functional data:

For each blood-oxygen level dependent (BOLD) run found per subject, the following preprocessing was performed. First, a reference volume and its skull-stripped version were generated for head motion correction. Head-motion parameters with respect to the BOLD reference (transformation matrices, and six corresponding rotation and translation parameters) were estimated before any spatiotemporal filtering using mcflirt (FSL 6.0.5.1; Jenkinson et al., 2002). The estimated fieldmap was then aligned with rigid-registration to the target EPI (echo-planar imaging) reference run and the field coefficients were mapped on to the reference EPI using the transform. BOLD runs were slice-time corrected to 0.346s (0.5 of slice acquisition range 0s-0.693s) using 3dTshift from AFNI (Cox & Hyde, 1997). The BOLD reference was then co-registered to the T1w reference using bbregister (FreeSurfer) which implements boundary-based registration (Greve & Fischl, 2009). This co-registration was configured with six degrees of freedom.

The BOLD time-series were resampled into standard space, generating a preprocessed BOLD run in MNI152NLin6Asym space. A reference volume and its skull-stripped version were generated using a custom methodology within fMRIPrep. The BOLD time-series were resampled onto the fsaverage surface (FreeSurfer). Grayordinates files (Glasser et al., 2013) containing 91k samples were also generated using the highest-resolution fsaverage as intermediate standardized surface space. All resamplings can be performed with a single interpolation step by composing all the pertinent transformations (i.e. head-motion transform matrices, susceptibility distortion correction when available, and co-registrations to anatomical and output spaces). Gridded (volumetric) resamplings were performed using antsApplyTransforms (ANTs), configured with Lanczos interpolation to minimize the smoothing effects of other kernels (Lanczos, 1964). Non-gridded (surface) resamplings were performed using mri_vol2surf (FreeSurfer).

Developmental sample postprocessing (HCPD and HBN):

Preprocessed outputs were post-processed using the eXtensible Connectivity Pipelines–DCAN collaborative (XCP-D v0.3.0; Ciric et al., 2018; Mehta et al., 2024).

Specifically, before nuisance regression, the BOLD data were despiked, mean-centered, and linearly detrended. In total, 36 nuisance regressors were selected from the nuisance confound matrices of fMRIPrep output. These regressors included six motion parameters, global signal, mean white-matter signal, and mean cerebrospinal fluid (CSF) signal with their temporal derivatives, as well as the quadratic expansion of six motion parameters, tissues signals and their temporal derivatives (Ciric et al., 2017; Satterthwaite et al., 2013). Nuisance regressors were regressed from the BOLD data using linear regression implemented in Scikit-Learn 1.1.3 (Pedregosa et al., 2011). The residual time series were then band-pass filtered to retain signals within the 0.01–0.08 Hz frequency band. Finally, processed functional time series were extracted using Connectome Workbench (Glasser et al., 2013) for the Schaefer-400 atlas (Schaefer et al., 2018), yielding parcellated time series for 400 cortical regions. A lower-resolution Schaefer-200 parcellation was also used for sensitivity analyses.

Young adult sample (HCPYA):

All HCPYA fMRI data were preprocessed using the HCP Minimal Preprocessing Pipelines (Glasser et al., 2013), which include the same major steps as above: gradient distortion correction, field-map distortion correction, boundary-based registration, bias-field correction, and whole-brain intensity normalization.

The minimally preprocessed HCP outputs were then post-processed using XCP-D v0.9.1, following a workflow closely analogous to that used for the developmental sample. Framewise displacement was computed using the same formula (Power et al., 2014), and 36 nuisance regressors (“36P”) were selected following the same definitions as above (Ciric et al., 2017; Satterthwaite et al., 2013). The BOLD data were converted to NIfTI format, despiked with AFNI’s 3dDespike, and converted back to CIFTI format.

Denoising was performed using a regression approach based on Nilearn’s implementation. Time series and confounds were band-pass filtered (0.01–0.08 Hz) using a second-order Butterworth filter, and nuisance regression was applied to the filtered confounds. The denoised BOLD data were then used to extract parcellated time series for the Schaefer-400 atlas using Connectome Workbench. As with the developmental sample, a Schaefer-200 parcellation was also generated for sensitivity analyses.

Intrinsic timescale

Fully processed parcellated fMRI time series were used to estimate intrinsic timescale for each brain region (i.e., parcel) and participant. Specifically, we first normalized the fMRI time series at each parcel using a z-score normalization and estimated an autocorrelation function (ACF) for the normalized time series using the Time Series Analysis (tsa) tool from statsmodels (Seabold & Perktold, 2010) in Python. Next, we identified the time point where ACF crosses zero autocorrelation value for the first time (i.e., first zero-crossing time point) and calculated the sum of all positive ACF values up to the first zero-crossing point. To obtain an estimate of the intrinsic timescale in seconds, we multiplied the sum of positive ACF values by the corresponding TR for each dataset (HCPD TR=800ms; HBN TR=800ms; HCPYA TR=720ms; Fulcher et al., 2013; Fulcher & Jones, 2017; Watanabe et al., 2019). Note that the fMRI acquisition TR corresponds to the time step between consecutive time points in ACF. This procedure is highly similar to calculating the area under the curve of the positive part of the ACF curve (Figure 1). We repeated this procedure for each parcel and individual, generating a parcellated map of intrinsic timescale for each participant.

We restricted our analyses to low-motion individuals (mean FD < 0.2 mm; see “Participants” for details), given that measures of intrinsic timescale are sensitive to in-scanner motion and physiological noise and are more reliable in higher-quality data with low motion (Goldberg et al., 2024). In addition, we included data quality (i.e., mean FD) as a covariate in subsequent models to account for any residual confounding effects of motion (see “Generalized Additive Model (GAM)” for details).

Generalized Additive Model (GAM)

We used Generalized Additive Models (GAMs) to assess linear and nonlinear associations with age in fMRI intrinsic timescale (Pomponio et al., 2020; Sydnor et al., 2023; Luo et al, 2024). The analysis was performed using the “mgcv” package in R 4.2.2. We included age as a smooth term and sex and data quality (i.e., in-scanner motion quantified by mean FD) as linear covariates in each GAM. We used Restricted Maximum Likelihood (REML) to estimate the smoothing parameters and set the maximum basis complexity to 3 for the smooth age term to avoid overfitting. Specifically, GAMs were formulated as `mgcv::gam(feature ~ s(age, k=3, fx=F) + factor(sex) + data_quality)`. We performed two sets of analysis: (1) whole-brain GAMs, where we assessed developmental patterns of the whole-brain average intrinsic timescale; (2) regional GAMs, where we examined developmental patterns of regional intrinsic timescale.

The effect size of associations between intrinsic timescale and age were estimated as partial R2. Partial R2 was quantified as the difference in Sum of Squared Errors (SSE) between the full model that included the smooth age term and a reduced model with no age term (i.e., only including model covariates) normalized by SSE of the reduced model. Normalization by SSE of the reduced model highlights the relative contribution of the predictor (e.g., age) to the reduced model. We used a signed version of partial R2, such that the sign reflected the directionality of observed effects (increase or decrease in intrinsic timescale with increasing age). To obtain the directionality of partial R2, we calculated the mean derivative of the model fit for the smooth term (i.e., age). A positive sign was assigned to partial R2 if the mean derivative was positive, reflecting an overall increasing trend between the intrinsic timescale and age, whereas a negative sign was assigned to partial R2 for negative mean derivative. The statistical significance of the associations between intrinsic timescale and age was assessed using analysis of variance (ANOVA) between the full model and reduced model that excluded the smooth age term. Results were corrected for multiple comparisons by controlling for the false discovery rate (FDR correction; Q<0.05).

Spatial null model

We used spatial autocorrelation-preserving null models (i.e., “spin” tests) to compare cortical patterns between different features or datasets (Alexander-Bloch et al., 2018; Markello et al., 2021). Specifically, we generated null brain maps with a randomized spatial distribution while preserving the spatial autocorrelation inherent to the data using the neuromaps toolbox (Markello et al., 2022). The randomization was implemented by applying 10,000 randomly sampled rotations to the spherical projections of the data for the Schaefer atlas (Alexander-Bloch et al., 2018; Markello et al., 2022). The rotations were applied to one hemisphere and then mirrored to the other hemisphere.

The resulting randomized (“spun”) brain maps were used for two types of significance testing. First, to evaluate the significance of associations between two cortical maps (e.g., the correlation between age effects and the S–A axis), we generated a null distribution of association statistics (e.g., Spearman’s rank correlation coefficient) by correlating one empirical map with 10,000 spun versions of the other map. The statistical significance of the observed association was determined by comparing the empirical statistic to this null distribution using a two-tailed test. Second, for analyses testing differences in mean values between two distributions (e.g., t-tests), spun brain maps were used to generate a null distribution of mean differences. The empirical mean difference was then compared against this null distribution to assess statistical significance (two-tailed test).

Supplementary Material

Supplement 1

Funding

Support was provided by grants from the National Institutes of Mental Health, including 2R01MH113550, 2R01MH112847, R37MH125829, R01EB022573, and U24NS130411. GS was supported by a postdoctoral fellowship from the Canadian Institutes of Health Research (CIHR).

Footnotes

Competing interests

The authors have declared no competing interest.

Code and data availability

All code and accompanying guidelines used to conduct the reported analyses are available on GitHub (https://pennlinc.github.io/shafiei_timescale). Data used in the present study were obtained from publicly available HCPYA (van Essen et al., 2013; https://www.humanconnectome.org/study/hcp-young-adult), HCPD (Somerville et al., 2018; https://www.humanconnectome.org/study/hcp-lifespan-development), and HBN (Alexander et al., 2017; https://fcon_1000.projects.nitrc.org/indi/cmi_healthy_brain_network/) datasets.

References

  1. Alexander L. M., Escalera J., Ai L., Andreotti C., Febre K., Mangone A., Vega-Potler N., Langer N., Alexander A., Kovacs M., Litke S., O’Hagan B., Andersen J., Bronstein B., Bui A., Bushey M., Butler H., Castagna V., Camacho N., … Milham M. P. (2017). An open resource for transdiagnostic research in pediatric mental health and learning disorders. Scientific Data, 4(1). 10.1038/sdata.2017.181 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Alexander-Bloch A. F., Shou H., Liu S., Satterthwaite T. D., Glahn D. C., Shinohara R. T., Vandekar S. N., & Raznahan A. (2018). On testing for spatial correspondence between maps of human brain structure and function. NeuroImage, 178, 540–551. 10.1016/j.neuroimage.2018.05.070 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Avants B., Epstein C., Grossman M., & Gee J. (2008). Symmetric diffeomorphic image registration with cross-correlation: Evaluating automated labeling of elderly and neurodegenerative brain. Medical Image Analysis, 12(1), 26–41. 10.1016/j.media.2007.06.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Bero J., Humphries C., Li Y., Kumar A., Lee H., Shinn M., Murray J. D., Vickery T. J., & Lee D. (2026). Temporal and spatial scales of human resting-state cortical activity across the lifespan. The Journal of Neuroscience, 46(2), e0577252025. 10.1523/jneurosci.0577-25.2025 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Beul S. F., & Hilgetag C. C. (2019). Neuron density fundamentally relates to architecture and connectivity of the primate cerebral cortex. NeuroImage, 189, 777–792. 10.1016/j.neuroimage.2019.01.010 [DOI] [PubMed] [Google Scholar]
  6. Burt J. B., Demirtaş M., Eckner W. J., Navejar N. M., Ji J. L., Martin W. J., Bernacchia A., Anticevic A., & Murray J. D. (2018). Hierarchy of transcriptomic specialization across human cortex captured by structural neuroimaging topography. Nature Neuroscience, 21(9), 1251–1259. 10.1038/s41593-018-0195-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Chaudhuri R., Knoblauch K., Gariel M.-A., Kennedy H., & Wang X.-J. (2015). A large-scale circuit mechanism for hierarchical dynamical processing in the primate cortex. Neuron, 88(2), 419–431. 10.1016/j.neuron.2015.09.008 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Ciric R., Rosen A. F. G., Erus G., Cieslak M., Adebimpe A., Cook P. A., Bassett D. S., Davatzikos C., Wolf D. H., & Satterthwaite T. D. (2018). Mitigating head motion artifact in functional connectivity MRI. Nature Protocols, 13(12), 2801–2826. 10.1038/s41596-018-0065-y [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Ciric R., Wolf D. H., Power J. D., Roalf D. R., Baum G. L., Ruparel K., Shinohara R. T., Elliott M. A., Eickhoff S. B., Davatzikos C., Gur R. C., Gur R. E., Bassett D. S., & Satterthwaite T. D. (2017). Benchmarking of participant-level confound regression strategies for the control of motion artifact in studies of functional connectivity. NeuroImage, 154, 174–187. 10.1016/j.neuroimage.2017.03.020 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Cox R. W., & Hyde J. S. (1997). Software tools for analysis and visualization of fMRI data. NMR in Biomedicine, 10(4–5), 171–178. 10.1002/(sici)1099-1492(199706/08)10:4/5<171::aid-nbm453>3.0.co;2-l [DOI] [PubMed] [Google Scholar]
  11. Dale A. M., Fischl B., & Sereno M. I. (1999). Cortical surface-based analysis. NeuroImage, 9(2), 179–194. 10.1006/nimg.1998.0395 [DOI] [PubMed] [Google Scholar]
  12. Demirtaş M., Burt J. B., Helmer M., Ji J. L., Adkinson B. D., Glasser M. F., Van Essen D. C., Sotiropoulos S. N., Anticevic A., & Murray J. D. (2019). Hierarchical heterogeneity across human cortex shapes large-scale neural dynamics. Neuron, 101(6), 1181–1194.e13. 10.1016/j.neuron.2019.01.017 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Esteban O., Markiewicz C. J., Blair R. W., Moodie C. A., Isik A. I., Erramuzpe A., Kent J. D., Goncalves M., DuPre E., Snyder M., Oya H., Ghosh S. S., Wright J., Durnez J., Poldrack R. A., & Gorgolewski K. J. (2018). fMRIPrep: A robust preprocessing pipeline for functional MRI. Nature Methods, 16(1), 111–116. 10.1038/s41592-018-0235-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Fulcher B. D., & Jones N. S. (2017). hctsa : A Computational Framework for Automated Time-Series Phenotyping Using Massive Feature Extraction. Cell Systems, 5(5), 527–531.e3. 10.1016/j.cels.2017.10.001 [DOI] [PubMed] [Google Scholar]
  15. Fulcher B. D., Little M. A., & Jones N. S. (2013). Highly comparative time-series analysis: The empirical structure of time series and their methods. Journal of The Royal Society Interface, 10(83), 20130048. 10.1098/rsif.2013.0048 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Gao R., van den Brink R. L., Pfeffer T., & Voytek B. (2020). Neuronal timescales are functionally dynamic and shaped by cortical microarchitecture. eLife, 9. 10.7554/elife.61277 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Glasser M. F., Sotiropoulos S. N., Wilson J. A., Coalson T. S., Fischl B., Andersson J. L., Xu J., Jbabdi S., Webster M., Polimeni J. R., Van Essen D. C., & Jenkinson M. (2013). The minimal preprocessing pipelines for the Human Connectome Project. NeuroImage, 80, 105–124. 10.1016/j.neuroimage.2013.04.127 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Goldberg A., Rosario I., Power J., Horga G., & Wengler K. (2024). Strategies for motion- and respiration-robust estimation of fMRI intrinsic neural timescales. Imaging Neuroscience, 2. 10.1162/imag_a_00326 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Golesorkhi M., Gomez-Pilar J., Tumati S., Fraser M., & Northoff G. (2021a). Temporal hierarchy of intrinsic neural timescales converges with spatial core-periphery organization. Communications Biology, 4(1). 10.1038/s42003-021-01785-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Golesorkhi M., Gomez-Pilar J., Zilio F., Berberian N., Wolff A., Yagoub M. C. E., & Northoff G. (2021b). The brain and its time: Intrinsic neural timescales are key for input processing. Communications Biology, 4(1). 10.1038/s42003-021-02483-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Gorgolewski K., Burns C. D., Madison C., Clark D., Halchenko Y. O., Waskom M. L., & Ghosh S. S. (2011). Nipype: A flexible, lightweight and extensible neuroimaging data processing framework in Python. Frontiers in Neuroinformatics, 5. 10.3389/fninf.2011.00013 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Greve D. N., & Fischl B. (2009). Accurate and robust brain image alignment using boundary-based registration. NeuroImage, 48(1), 63–72. 10.1016/j.neuroimage.2009.06.060 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Hasson U., Yang E., Vallines I., Heeger D. J., & Rubin N. (2008). A hierarchy of temporal receptive windows in human cortex. The Journal of Neuroscience, 28(10), 2539–2550. 10.1523/jneurosci.5487-07.2008 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Honey C. J., Thesen T., Donner T. H., Silbert L. J., Carlson C. E., Devinsky O., Doyle W. K., Rubin N., Heeger D. J., & Hasson U. (2012). Slow cortical dynamics and the accumulation of information over long timescales. Neuron, 76(2), 423–434. 10.1016/j.neuron.2012.08.011 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Huntenburg J. M., Bazin P.-L., & Margulies D. S. (2018). Large-scale gradients in human cortical organization. Trends in Cognitive Sciences, 22(1), 21–31. 10.1016/j.tics.2017.11.002 [DOI] [PubMed] [Google Scholar]
  26. Ito T., Hearne L. J., & Cole M. W. (2020). A cortical hierarchy of localized and distributed processes revealed via dissociation of task activations, connectivity changes, and intrinsic timescales. NeuroImage, 221, 117141. 10.1016/j.neuroimage.2020.117141 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Kauvar I., Richman E. B., Liu T. X., Li C., Vesuna S., Chibukhchyan A., Yamada L., Fogarty A., Solomon E., Choi E. Y., Mortazavi L., Chau Loo Kung G., Mukunda P., Raja C., Gil-Hernández D., Patron K., Zhang X., Brawer J., Wrobel S., … Deisseroth K. (2025). Conserved brain-wide emergence of emotional response from sensory experience in humans and mice. Science, 388(6750). 10.1126/science.adt3971 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Kiebel S. J., Daunizeau J., & Friston K. J. (2008). A hierarchy of time-scales and the brain. PLoS Computational Biology, 4(11), e1000209. 10.1371/journal.pcbi.1000209 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Lanczos C. (1964). Evaluation of noisy data. Journal of the Society for Industrial and Applied Mathematics Series B Numerical Analysis, 1(1), 76–85. 10.1137/0701007 [DOI] [Google Scholar]
  30. Li X., Bianchini Esper N., Ai L., Giavasis S., Jin H., Feczko E., Xu T., Clucas J., Franco A., Sólon Heinsfeld A., Adebimpe A., Vogelstein J. T., Yan C.-G., Esteban O., Poldrack R. A., Craddock C., Fair D., Satterthwaite T., Kiar G., & Milham M. P. (2024). Moving beyond processing- and analysis-related variation in resting-state functional brain imaging. Nature Human Behaviour, 8(10), 2003–2017. 10.1038/s41562-024-01942-4 [DOI] [PubMed] [Google Scholar]
  31. Luo A. C., Sydnor V. J., Pines A., Larsen B., Alexander-Bloch A. F., Cieslak M., Covitz S., Chen A. A., Esper N. B., Feczko E., Franco A. R., Gur R. E., Gur R. C., Houghton A., Hu F., Keller A. S., Kiar G., Mehta K., Salum G. A., … Satterthwaite T. D. (2024). Functional connectivity development along the sensorimotor-association axis enhances the cortical hierarchy. Nature Communications, 15(1). 10.1038/s41467-024-47748-w [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Margulies D. S., Ghosh S. S., Goulas A., Falkiewicz M., Huntenburg J. M., Langs G., Bezgin G., Eickhoff S. B., Castellanos F. X., Petrides M., Jefferies E., & Smallwood J. (2016). Situating the default-mode network along a principal gradient of macroscale cortical organization. Proceedings of the National Academy of Sciences, 113(44), 12574–12579. 10.1073/pnas.1608282113 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Markello R. D., Hansen J. Y., Liu Z.-Q., Bazinet V., Shafiei G., Suárez L. E., Blostein N., Seidlitz J., Baillet S., Satterthwaite T. D., Chakravarty M. M., Raznahan A., & Misic B. (2022). neuromaps: Structural and functional interpretation of brain maps. Nature Methods, 19(11), 1472–1479. 10.1038/s41592-022-01625-w [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Markello R. D., & Misic B. (2021). Comparing spatial null models for brain maps. NeuroImage, 236, 118052. 10.1016/j.neuroimage.2021.118052 [DOI] [PubMed] [Google Scholar]
  35. McKeon S. D., Petrie D., Sydnor V. J., Wang Z., Zhu J., Famalette A., Foran W., Parr A. C., Calabro F. J., Abel T., Constantinidis C., & Luna B. (2025). Cross-species evidence for the refinement of intrinsic neural timescales supporting executive system maturation through adolescence. openRxiv. 10.64898/2025.12.12.694027 [DOI] [Google Scholar]
  36. Mehta K., Salo T., Madison T. J., Adebimpe A., Bassett D. S., Bertolero M., Cieslak M., Covitz S., Houghton A., Keller A. S., Lundquist J. T., Luo A., Miranda-Dominguez O., Nelson S. M., Shafiei G., Shanmugan S., Shinohara R. T., Smyser C. D., Sydnor V. J., … Satterthwaite T. D. (2024). XCP-D: A robust pipeline for the post-processing of fMRI data. Imaging Neuroscience, 2. 10.1162/imag_a_00257 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Miles J. T., Weaver K. E., Webb S. J., & Ojemann J. G. (2026). Developmental relationships between the human alpha rhythm and intrinsic neural timescales are dependent on neural hierarchy. Journal of Neurophysiology, 135(1), 143–152. 10.1152/jn.00435.2025 [DOI] [PubMed] [Google Scholar]
  38. Murray J. D., Bernacchia A., Freedman D. J., Romo R., Wallis J. D., Cai X., Padoa-Schioppa C., Pasternak T., Seo H., Lee D., & Wang X.-J. (2014). A hierarchy of intrinsic timescales across primate cortex. Nature Neuroscience, 17(12), 1661–1663. 10.1038/nn.3862 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Murray J. D., Demirtaş M., & Anticevic A. (2018). Biophysical modeling of large-scale brain dynamics and applications for computational psychiatry. Biological Psychiatry: Cognitive Neuroscience and Neuroimaging, 3(9), 777–787. 10.1016/j.bpsc.2018.07.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Nishio M., Ellwood-Lowe M. E., Woodburn M., McDermott C. L., Park A. T., Tooley U. A., Boroshok A. L., Grandjean J., & Mackey A. P. (2024). The Hurst exponent as a marker of inhibition in the developing brain. openRxiv. 10.1101/2024.09.29.615675 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Pedregosa F., Varoquaux G., Gramfort A., Michel V., Thirion B., Grisel O., ... & Duchesnay É. (2011). Scikit-learn: Machine learning in Python. The Journal of Machine Learning Research, 12, 2825–2830. [Google Scholar]
  42. Pomponio R., Erus G., Habes M., Doshi J., Srinivasan D., Mamourian E., Bashyam V., Nasrallah I. M., Satterthwaite T. D., Fan Y., Launer L. J., Masters C. L., Maruff P., Zhuo C., Völzke H., Johnson S. C., Fripp J., Koutsouleris N., Wolf D. H., … Davatzikos C. (2020). Harmonization of large MRI datasets for the analysis of brain imaging patterns throughout the lifespan. NeuroImage, 208, 116450. 10.1016/j.neuroimage.2019.116450 [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Power J. D., Mitra A., Laumann T. O., Snyder A. Z., Schlaggar B. L., & Petersen S. E. (2014). Methods to detect, characterize, and remove motion artifact in resting state fMRI. NeuroImage, 84, 320–341. 10.1016/j.neuroimage.2013.08.048 [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Raut R. V., Snyder A. Z., & Raichle M. E. (2020). Hierarchical dynamics as a macroscopic organizing principle of the human brain. Proceedings of the National Academy of Sciences, 117(34), 20890–20897. 10.1073/pnas.2003383117 [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Sadaghiani S., & Wirsich J. (2020). Intrinsic connectome organization across temporal scales: New insights from cross-modal approaches. Network Neuroscience, 4(1), 1–29. 10.1162/netn_a_00114 [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Satterthwaite T. D., Elliott M. A., Gerraty R. T., Ruparel K., Loughead J., Calkins M. E., Eickhoff S. B., Hakonarson H., Gur R. C., Gur R. E., & Wolf D. H. (2013). An improved framework for confound regression and filtering for control of motion artifact in the preprocessing of resting-state functional connectivity data. NeuroImage, 64, 240–256. 10.1016/j.neuroimage.2012.08.052 [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Schaefer A., Kong R., Gordon E. M., Laumann T. O., Zuo X.-N., Holmes A. J., Eickhoff S. B., & Yeo B. T. T. (2018). Local-global parcellation of the human cerebral cortex from intrinsic functional connectivity MRI. Cerebral Cortex, 28(9), 3095–3114. 10.1093/cercor/bhx179 [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Scheeringa R., Koopmans P. J., van Mourik T., Jensen O., & Norris D. G. (2016). The relationship between oscillatory EEG activity and the laminar-specific BOLD signal. Proceedings of the National Academy of Sciences, 113(24), 6761–6766. 10.1073/pnas.1522577113 [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Seabold S., & Perktold J. (2010). Statsmodels: Econometric and statistical modeling with Python. Proceedings of the Python in Science Conference, 92–96. 10.25080/majora-92bf1922-011 [DOI] [Google Scholar]
  50. Shafiei G., Esper N., Hoffmann M., Ai L., Chen A., Cluce J., Covitz S., Giavasis S., Lane C., Mehta K., Moore T., Salo T., Tapera T., Calkins M., Colcombe S., Davatzikos C., Gur R., Gur R., Pan P., … Satterthwaite T. (2025). Reproducible Brain Charts: An open data resource for mapping brain development and its associations with mental health. openRxiv. 10.1101/2025.02.24.639850 [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Shafiei G., Fulcher B. D., Voytek B., Satterthwaite T. D., Baillet S., & Misic B. (2023). Neurophysiological signatures of cortical micro-architecture. Nature Communications, 14(1). 10.1038/s41467-023-41689-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Shafiei G., Markello R. D., Vos de Wael R., Bernhardt B. C., Fulcher B. D., & Misic B. (2020). Topographic gradients of intrinsic dynamics across neocortex. eLife, 9. 10.7554/elife.62116 [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Somerville L. H., Bookheimer S. Y., Buckner R. L., Burgess G. C., Curtiss S. W., Dapretto M., Elam J. S., Gaffrey M. S., Harms M. P., Hodge C., Kandala S., Kastman E. K., Nichols T. E., Schlaggar B. L., Smith S. M., Thomas K. M., Yacoub E., Van Essen D. C., & Barch D. M. (2018). The Lifespan Human Connectome Project in Development: A large-scale study of brain connectivity development in 5–21 year olds. NeuroImage, 183, 456–468. 10.1016/j.neuroimage.2018.08.050 [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Sydnor V. J., Larsen B., Bassett D. S., Alexander-Bloch A., Fair D. A., Liston C., Mackey A. P., Milham M. P., Pines A., Roalf D. R., Seidlitz J., Xu T., Raznahan A., & Satterthwaite T. D. (2021). Neurodevelopment of the association cortices: Patterns, mechanisms, and implications for psychopathology. Neuron, 109(18), 2820–2846. 10.1016/j.neuron.2021.06.016 [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Sydnor V. J., Larsen B., Seidlitz J., Adebimpe A., Alexander-Bloch A. F., Bassett D. S., Bertolero M. A., Cieslak M., Covitz S., Fan Y., Gur R. E., Gur R. C., Mackey A. P., Moore T. M., Roalf D. R., Shinohara R. T., & Satterthwaite T. D. (2023). Intrinsic activity development unfolds along a sensorimotor–association cortical axis in youth. Nature Neuroscience, 26(4), 638–649. 10.1038/s41593-023-01282-y [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Truzzi A., & Cusack R. (2023). The development of intrinsic timescales: A comparison between the neonate and adult brain. NeuroImage, 275, 120155. 10.1016/j.neuroimage.2023.120155 [DOI] [PubMed] [Google Scholar]
  57. Tustison N. J., Avants B. B., Cook P. A., Zheng Yuanjie, Egan A., Yushkevich P. A., & Gee J. C. (2010). N4ITK: Improved N3 bias correction. IEEE Transactions on Medical Imaging, 29(6), 1310–1320. 10.1109/tmi.2010.2046908 [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Van Essen D. C., Smith S. M., Barch D. M., Behrens T. E. J., Yacoub E., & Ugurbil K. (2013). The WU-Minn Human Connectome Project: An overview. NeuroImage, 80, 62–79. 10.1016/j.neuroimage.2013.05.041 [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Wang X.-J. (1999). Synaptic basis of cortical persistent activity: The importance of NMDA receptors to working memory. The Journal of Neuroscience, 19(21), 9587–9603. 10.1523/jneurosci.19-21-09587.1999 [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Wang X.-J. (2002). Probabilistic decision making by slow reverberation in cortical circuits. Neuron, 36(5), 955–968. 10.1016/s0896-6273(02)01092-9 [DOI] [PubMed] [Google Scholar]
  61. Wang X.-J. (2008). Decision making in recurrent neuronal circuits. Neuron, 60(2), 215–234. 10.1016/j.neuron.2008.09.034 [DOI] [PMC free article] [PubMed] [Google Scholar]
  62. Wang X.-J. (2020). Macroscopic gradients of synaptic excitation and inhibition in the neocortex. Nature Reviews Neuroscience, 21(3), 169–178. 10.1038/s41583-020-0262-x [DOI] [PMC free article] [PubMed] [Google Scholar]
  63. Wang X.-J. (2022). Theory of the multiregional neocortex: Large-scale neural dynamics and distributed cognition. Annual Review of Neuroscience, 45(1), 533–560. 10.1146/annurev-neuro-110920-035434 [DOI] [PMC free article] [PubMed] [Google Scholar]
  64. Waschke L., Kloosterman N. A., Obleser J., & Garrett D. D. (2021). Behavior needs neural variability. Neuron, 109(5), 751–766. 10.1016/j.neuron.2021.01.023 [DOI] [PubMed] [Google Scholar]
  65. Watanabe T., Rees G., & Masuda N. (2019). Atypical intrinsic neural timescale in autism. eLife, 8. 10.7554/elife.42256 [DOI] [PMC free article] [PubMed] [Google Scholar]
  66. Wengler K., Goldberg A. T., Chahine G., & Horga G. (2020). Distinct hierarchical alterations of intrinsic neural timescales account for different manifestations of psychosis. eLife, 9. 10.7554/elife.56151 [DOI] [PMC free article] [PubMed] [Google Scholar]
  67. Wu K., & Gollo L. L. (2025). Mapping and modeling age-related changes in intrinsic neural timescales. Communications Biology, 8(1). 10.1038/s42003-025-07517-x [DOI] [PMC free article] [PubMed] [Google Scholar]
  68. Xie K., Royer J., Lariviere S., Rodriguez-Cruces R., de Wael R. V., Park B., Auer H., Tavakol S., DeKraker J., Abdallah C., Caciagli L., Bassett D. S., Bernasconi A., Bernasconi N., Frauscher B., Concha L., & Bernhardt B. C. (2023). Atypical intrinsic neural timescales in temporal lobe epilepsy. Epilepsia, 64(4), 998–1011. 10.1111/epi.17541 [DOI] [PubMed] [Google Scholar]
  69. Zeng D., Li Q., Li D., & Li S. (2024). Shifts in intrinsic neural timescale of hippocampus support the maturation of inhibitory control and working memory in youth. Medicine in Novel Technology and Devices, 22, 100302. 10.1016/j.medntd.2024.100302 [DOI] [Google Scholar]
  70. Zhang Y., Brady M., & Smith S. (2001). Segmentation of brain MR images through a hidden Markov random field model and the expectation-maximization algorithm. IEEE Transactions on Medical Imaging, 20(1), 45–57. 10.1109/42.906424 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplement 1
NIHPP2026.04.08.717312v1-supplement-1.pdf (2.2MB, pdf)

Data Availability Statement

All code and accompanying guidelines used to conduct the reported analyses are available on GitHub (https://pennlinc.github.io/shafiei_timescale). Data used in the present study were obtained from publicly available HCPYA (van Essen et al., 2013; https://www.humanconnectome.org/study/hcp-young-adult), HCPD (Somerville et al., 2018; https://www.humanconnectome.org/study/hcp-lifespan-development), and HBN (Alexander et al., 2017; https://fcon_1000.projects.nitrc.org/indi/cmi_healthy_brain_network/) datasets.

Articles from bioRxiv are provided here courtesy of Cold Spring Harbor Laboratory Preprints

RESOURCES