Unraveling reproducible dynamic states of individual brain functional parcellation

Abstract

Data-driven parcellations are widely used for exploring the functional organization of the brain, and also for reducing the high dimensionality of fMRI data. Despite the flurry of methods proposed in the literature, functional brain parcellations are not highly reproducible at the level of individual subjects, even with very long acquisitions. Some brain areas are also more difficult to parcellate than others, with association heteromodal cortices being the most challenging. An important limitation of classical parcellations is that they are static, i.e. they neglect dynamic reconfigurations of brain networks. In this paper, we proposed a new method to identify dynamic states of parcellations, which we hypothesized would improve reproducibility over static parcellation approaches. For a series of seed voxels in the brain, we applied a cluster analysis to regroup short (3 minutes) time windows into “states” with highly similar seed parcels. We splitted individual time series of the Midnight scan club sample into two independent sets of 2.5 hours (test and retest). We found that average within-state parcellations, called stability maps, were highly reproducible (over .9 test-retest spatial correlation in many instances) and subject specific (fingerprinting accuracy over 70 percent on average) between test and retest. Consistent with our hypothesis, seeds in heteromodal cortices (posterior and anterior cingulate) showed a richer repertoire of states than unimodal (visual) cortex. Taken together, our results indicate that static functional parcellations are incorrectly averaging well-defined and distinct dynamic states of brain parcellations. This work calls to revisit previous methods based on static parcellations, which includes the majority of published network analyses of fMRI data. Our method may, thus, impact how researchers model the rich interactions between brain networks in health and disease.

Systematic evaluation of fMRI data-processing pipelines for consistent functional connectomics

“…Recent gradient-based and eigenmode-based approaches provide alternative representations of the brain that are spatially extended, overlapping, and continuous rather than discrete (analogously to ICA), offering a complementary perspective on the constituent elements of the brain's functional organisation 32,[112][113][114][115][116] . It is also worth mentioning that, whatever the macroscopic units of brain function may be in terms of space, they need not be temporally invariant: future extensions of the present work may consider node definition schemes that allow for node boundaries to vary, and nodes themselves to merge or split dynamically in time, or as a function of task 27,[117][118][119][120] . Finally, future approaches may enrich nodes with biological annotations such as microstructure, chemoarchitecture, and heterogeneities reflecting additional biological properties [93][94][95]121,122 , thereby providing a path towards a more integrative neuroscience.…”

Identifying canonical and replicable multi‐scale intrinsic connectivity networks in 100k+ resting‐state fMRI datasets

“…A key step toward establishing rsfMRI as a prevalent clinical tool is the accurate estimation of corresponding functional patterns, that is, the identification of equivalent functional patterns across individuals and brain states in a way that captures both individual variations and inter‐subject correspondence. Accumulating evidence of spatial differences in functional patterns across individuals and even within individuals over time (Bhinge et al, 2019 ; Boukhdhir et al, 2021 ; Fan et al, 2021 ; Iraji, Deramus, et al, 2019 ; Iraji, Fu, et al, 2019 ; Iraji et al, 2020 ; Luo et al, 2021 ; Salehi et al, 2020 ; Wu et al, 2021 ) highlights the necessity of using data‐driven approaches instead of predefined (anatomical or functional) atlases in FC studies. However, several factors must be taken into account when using data‐driven approaches.…”

“…By using anatomically fixed regions, this category implicitly assumes the functional (connectivity) profile within each anatomically fixed region does not vary over time and is the same across individuals. However, many static and dynamic rsfMRI studies have challenged this strong assumption by identifying meaningful and replicable differences in the spatial patterns of functional entities both across subjects and within subjects over time (Boukhdhir et al, 2021 ; Erhardt et al, 2011 ; Iraji, Deramus, et al, 2019 ; Iraji, Fu, et al, 2019 ; Luo et al, 2021 ; Wang et al, 2015 ). The presence of within‐ and between‐subject spatial differences is further supported by task‐based fMRI findings showing that functional connectivity maps and spatial patterns of brain responses vary across individuals for a given task as well as within‐subject across mental states dictated by tasks (Calhoun et al, 2008 ; Krienen et al, 2014 ; Salehi et al, 2020 ; Sui et al, 2009 ; Wu et al, 2021 ).…”

“…By using anatomically fixed regions, this category implicitly assumes the functional (connectivity) profile within each anatomically fixed region does not vary over time and is the same across individuals. However, many static and dynamic rsfMRI studies have challenged this strong assumption by identifying meaningful and replicable differences in the spatial patterns of functional entities both across subjects and within subjects over time (Boukhdhir et al, 2021;Erhardt et al, 2011;Luo et al, 2021;Wang et al, 2015).…”

Spatial Dynamic Subspaces Encode Sex-Specific Schizophrenia Disruptions in Transient Network Overlap and its Links to Genetic Risk

“…Temporal dynamics refer to variations in the temporal patterns of functional units, commonly assessed through changes in second-order statistics. Spatial dynamics, on the other hand, refer to variations in the spatial distribution of a source over time [1,[9][10][11][12]. The continuous reconfiguration of coordinated intrinsic activities can result in changes in the spatial patterns of functional units over time.…”