pipette
ENEnglish

Multistage gradient-based whole-cortex modeling for joint inference of regional circuit heterogeneity and interregional structural connectivity

B. Yang, Z. Jin, X. Chen, D. Zhou, S. Li

Preprint

En palabras de los autores

Current non-invasive neuroimaging methods cannot directly resolve regional circuit heterogeneity and interregional structural connectivity, limiting mechanistic understanding of brain structure-function relationships and accurate whole-brain modeling. Here we introduce StaGIC (Staged Gradient-based Inference of Circuitry), a method that jointly infers regional circuit heterogeneity and model-constrained structural connectivity (SC) from resting-state functional connectivity (FC) and functional connectivity dynamics (FCD). Ablation and synthetic-recovery experiments show that staged optimization improves fits to functional data and ground-truth parameter recovery, with inferred SC remaining highly consistent across retained solutions from independent optimization runs. In human data, freely optimized regional parameters exhibit consistent relationships with cortical hierarchy across parcellations, while SC inference substantially improves FC reconstruction. In macaque and marmoset data, inferred SC shows stronger correlation with tracer-derived projections than diffusion MRI estimates, particularly for long-range connections. As a proof of concept in computational psychiatry, group-level models of major depressive disorder (MDD) reproduce MDD-associated FC differences and generate exploratory hypotheses about hierarchical differentiation and interregional projection organization. StaGIC thus provides a differentiable framework for inferring multiscale cortical organization from non-invasive functional imaging.

Resultado principalLimitación que admiten los autores

Apareció: jueves, 19 de noviembre. bioRxiv. Preprint, todavía sin revisión por pares.

DOI: 10.64898/2026.11.16.767192