sfmc-infant-brain-stochastic-modules

name	sfmc-infant-brain-stochastic-modules
description	Stochastic module-based methodology for robust probabilistic measurement of structural-functional module consistency (SFMC) in brain networks. Accounts for inter-individual variability and reveals stronger developmental reorganization than conventional coupling approaches. Use for infant brain development analysis, structure-function coupling studies, and brain network module analysis.
activation	structural-functional module consistency, SFMC, stochastic modules, infant brain development, Baby Connectome, brain network modules, developmental reorganization, structure-function coupling
tags	["neuroscience","brain-development","structural-functional","infant-brain","network-modules","probabilistic-methods"]
version	1.0.0
author	agent
arxiv_id	2606.19739

Structural-Functional Module Consistency via Stochastic Modules

Core Contribution

Introduces stochastic modules within brain networks for robust probabilistic measurement of structural-functional module consistency (SFMC) across subjects. Overcomes limitations of conventional structure-function coupling approaches by accounting for inter-individual variability.

Key Innovation

Stochastic Module Definition

A stochastic module represents the probability of a brain region being assigned to a group-level sub-network across subjects, characterized as an assignment probability for each brain region.

Advantages Over Conventional Methods

1. Robustly evaluates consistency between structural and functional modules whose population sizes are not necessarily the same
2. Accounts for inter-individual variability of modules for groups
3. Reveals more pronounced decline in structure-function coupling, indicating stronger developmental reorganization than conventional SC-FC coupling approaches

Key Findings (Baby Connectome Project Data)

Developmental Trajectory (0-5 years)

• SFMC decreases from 0 to 5 years old — indicating progressive decoupling of structure and function during development
• Greater in primary brain regions (visual areas) — structure-function coupling remains stronger in sensory regions
• Lower in advanced cognitive regions — attention, control, and default mode network regions show weaker structure-function coupling

Methodological Significance

• Reveals stronger developmental reorganization than conventional SC-FC coupling
• The decoupling pattern aligns with known developmental hierarchies: primary sensory → association cortices

Methodology

Stochastic Module Framework

1. Partition structural network into modules
2. Partition functional network into modules
3. For each brain region across subjects:
   → Compute assignment probability to each module
   → Characterize as stochastic module membership
4. Measure consistency between SC and FC module assignments
5. Account for inter-individual variability in module composition

Comparison with Conventional Approaches

• Conventional SC-FC coupling: Measures correlation between SC and FC edge weights
• Stochastic module approach: Measures probabilistic consistency of modular organization
• Key difference: Conventional approach averages over subjects; stochastic approach captures variability

Applications

1. Infant brain development — tracking SC-FC decoupling trajectories
2. Neurodevelopmental disorders — atypical SFMC trajectories as biomarkers
3. Brain network evolution — understanding how structure-function relationships change
4. Individual differences — quantifying variability in modular organization
5. Hierarchical development — primary vs. association cortex maturation patterns

Pitfalls

1. Module size mismatch — SC and FC modules may have different sizes; method handles this
2. Inter-individual variability — conventional methods average out important variation
3. Developmental stage matters — SFMC trajectory is age-dependent
4. Region hierarchy — primary vs. association regions follow different trajectories
5. Population specificity — findings from BCP may not generalize to all populations

Related Skills

• structural-functional-brain-gnn — GNN-based structure-function learning
• linear-structure-function-coupling — linear SC-FC coupling framework
• brain-graph-neural — graph neural network brain connectivity analysis
• mesoscale-brain-organization — mesoscale structure identification
• weighted-brain-community-detection — weighted network community detection
• sfmc-structural-functional-module-consistency — related module consistency approach

Source

Bian, L., Liu, F., Wang, Q., Zhang, H., Shen, D., & the UNC/UMN Baby Connectome Project Consortium. (2026). Robust probabilistic measurement of structural-functional module consistency in infant brain development. arXiv:2606.19739. Published in Brain Structure and Function (DOI: 10.1007/s00429-026-03143-3).