Multi-Spatial Scale Dynamic Interactions between Functional Sources Reveal Sex-Specific Changes in Schizophrenia

Armin Iraji, Ashkan Faghiri, Zening Fu, Srinivas Rachakonda, Peter Kochunov, Ayşenil Belger, Judith M. Ford, Sarah McEwen, Daniel H. Mathalon, Bryon A. Mueller, Godfrey D. Pearlson, Steven G. Potkin, Adrian Preda, Jessica A. Turner, Theo G.m. van Erp, Vince D. CalhounbioRxiv 2021 · DOI

TL;DR

This paper introduces "multiscale ICA" (msICA), a novel method to analyze brain functional connectivity across different spatial scales simultaneously. Applying this to schizophrenia data, they found significant sex-specific differences in how brain regions interact, suggesting that a multiscale, sex-aware approach is crucial for understanding this disorder and developing targeted biomarkers.

What Problem Does This Paper Address?

Traditional brain functional connectivity (FC) analysis often relies on predefined anatomical regions or a single "model order" for Independent Component Analysis (ICA), which can oversimplify the brain's complex organization. This approach can miss interactions occurring at different spatial scales and ignore how these scales interact with each other. Furthermore, while sex differences are known to exist in schizophrenia, they are often understudied, especially in the context of dynamic, multiscale brain interactions. This paper aims to bridge this gap by developing a method that can capture these multiscale interactions and investigate sex-specific alterations in schizophrenia.

Methods

The core innovation is multiscale ICA (msICA). This method extends traditional ICA by running it with multiple "model orders" (representing different levels of spatial granularity) simultaneously. This allows msICA to identify functional sources (Intrinsic Connectivity Networks or ICNs) at various scales without pre-defining their size or location.

Once these multiscale ICNs are identified, the researchers use Functional Network Connectivity (FNC) to analyze the interactions between them. They employ both static FNC (sFNC), which averages connectivity over the entire scan, and dynamic FNC (dFNC), which captures how connectivity changes over time by analyzing short time windows. dFNC is further analyzed using k-means clustering to identify distinct "dynamic states" of brain connectivity.

Finally, they conduct group comparison analyses to identify sex-specific differences in sFNC and dFNC between individuals with schizophrenia (SZ) and healthy controls (CT), comparing males to males and females to females, and then comparing these sex-specific differences.

Datasets

The study utilized resting-state fMRI (rsfMRI) data from 827 individuals, comprising 477 typical controls and 350 individuals with schizophrenia. This data was pooled from three projects:

  • FBIRN (Functional Imaging Biomedical Informatics Research Network)
  • MPRC (Maryland Psychiatric Research Center)
  • COBRE (Center for Biomedical Research Excellence)

The paper does not introduce new benchmarks.

Key Insights and Ideas

  • Brain connectivity isn't monolithic: The brain's functional organization exists at multiple spatial scales, and these scales interact. Ignoring this multiscale nature can lead to an incomplete understanding of brain function and dysfunction.
  • msICA offers a richer view: By analyzing connectivity across different granularities simultaneously, msICA provides a more comprehensive picture than single-scale approaches. It avoids forcing functional sources into predefined sizes, allowing the data to reveal natural organizational patterns.
  • Schizophrenia shows sex-specific disruptions: The study found that the way brain regions connect differently in schizophrenia is not uniform across sexes. These sex-specific differences are particularly evident in dynamic connectivity patterns.
  • Dynamic states matter: Different "states" of brain connectivity emerge over time, and these states show distinct sex-specific alterations in schizophrenia. This highlights the importance of looking beyond static connectivity.
  • Potential for biomarkers: The observed sex-specific functional connectivity changes, especially those correlating with symptom severity, suggest that a sex-aware approach is crucial for identifying reliable biomarkers for schizophrenia.

Key Results

  • msICA successfully identified ICNs at multiple spatial scales, with higher model orders generally leading to more granular segregation of brain regions.
  • sFNC analysis revealed similar patterns across groups, but dFNC analysis identified four distinct dynamic states of brain connectivity.
  • Significant sex-specific differences in FNC were observed, particularly in dynamic states.
  • Males with schizophrenia showed larger alterations in several FNC pairs compared to females with schizophrenia.
  • Specific domains like the subcortical, cerebellar, temporal, and motor domains were highlighted as showing prominent sex-dependent alterations.
  • dFNC State 4 showed the most pronounced sex-specific differences, with some connections showing opposite patterns of alteration between males and females.
  • Sex-specific differences in functional domain connectivity correlated with symptom scores in males for certain features, but not in females.

Limitations and Future Work

  • Limited range of model orders: The study used only four model orders (25, 50, 75, 100). Future work should explore a wider range to capture more spatial levels of functional hierarchy.
  • Back-reconstruction approaches: Different back-reconstruction techniques for ICA could be explored for their impact on results.
  • Temporal dynamics of spatial patterns: While dynamic FNC was analyzed, future studies could investigate how the spatial patterns of functional sources themselves change over time.
  • Unbalanced sample sizes: The study acknowledges an unbalanced number of females and males, particularly in the schizophrenia group. Future studies should aim for larger datasets with more balanced sex representation.
  • Medication and treatment effects: Long-term effects of medication and treatment were not fully accounted for and could influence observed differences.
  • Longitudinal studies: Future longitudinal studies are recommended to understand how sex differences in schizophrenia evolve over time and with age of onset.
  • At-risk populations: Investigating at-risk unmedicated individuals could provide further insights into brain function changes.
  • Broader application of msICA: The authors advocate for the broader application of msICA and similar multiscale methods in future research.

Summary updated

Abstract

Abstract We introduce an extension of independent component analysis (ICA), called multiscale ICA (msICA), and design an approach to capture dynamic functional source interactions within and between multiple spatial scales. msICA estimates functional sources at multiple spatial scales without imposing direct constraints on the size of functional sources, overcomes the limitation of using fixed anatomical locations, and eliminates the need for model-order selection in ICA analysis. We leveraged this approach to study sex-specific and -common connectivity patterns in schizophrenia. Results show dynamic reconfiguration and interaction within and between multi-spatial scales. Sex-specific differences occur (1) within the subcortical domain, (2) between the somatomotor and cerebellum domains, and (3) between the temporal domain and several others, including the subcortical, visual, and default mode domains. Most of the sex-specific differences belong to between-spatial scale functional interactions and are associated with a dynamic state with strong functional interactions between the visual, somatomotor, and temporal domains and their anticorrelation patterns with the rest of the brain. We observed significant correlations between multi-spatial scale functional interactions and symptom scores, highlighting the importance of multiscale analyses to identify potential biomarkers for schizophrenia. As such, we recommend such analyses as an important option for future functional connectivity studies.
BibTeX
@misc{iraji2021multi,
  author = {Armin Iraji and Ashkan Faghiri and Zening Fu and Srinivas Rachakonda and Peter Kochunov and Ayşenil Belger and Judith M. Ford and Sarah McEwen and Daniel H. Mathalon and Bryon A. Mueller and Godfrey D. Pearlson and Steven G. Potkin and Adrian Preda and Jessica A. Turner and Theo G.M. van Erp and Vince D. Calhoun},
  title = {{Multi-Spatial Scale Dynamic Interactions between Functional Sources Reveal Sex-Specific Changes in Schizophrenia}},
  year = {2021},
  doi = {10.1101/2021.01.04.425222},
  howpublished = {bioRxiv},
}
Figure