Infraslow generation and transmission shape BOLD connectivity in neural mass models

Resting-state functional connectivity (FC) is measured from slow blood-oxygen-level-dependent (BOLD) fluctuations, but their origin and its consequences for FC remain unresolved. We compared two sources of infraslow activity (ISO; 0.01–0.1 Hz) in pyramidal–interneuron gamma (PING)-like neural-mass networks: autonomous gamma interactions in noise-free networks (PING⁺) and regionally independent external noise in locally damped networks (PING⁻). Group-average Human Connectome Project diffusion-MRI structural connectivity (SC) and resting-state FC, distributed through the ENIGMA Toolbox, supplied the 82-region network and empirical benchmark. Frequency-selective interventions, calibrated Stuart–Landau (SL) comparators and BOLD decomposition separated generation from transmission. In PING⁺, carrier interventions and temporal prediction supported ISO generation by gamma beating. Converting beating between nearby gamma rhythms into a slow rate component requires even-order mixing, supplied here by the sigmoid and omitted from the first-order SL comparison. This endogenous source is constrained by SC, circuit parameters and tonic inputs; the coupled-equilibrium linearization contained no slow modes. In PING⁻, noise prescribes the ISO input and neural ISO persists without gamma exchange. Both regimes produced SC-aligned FC, but their FC patterns differed, and PING⁺ showed stronger empirical correspondence at the tested operating points and record lengths. With first-order transmitted rates and readout, noise-free SL⁺ produced negligible ISO, whereas noise-driven SL⁻ retained ISO and structural alignment. Suppressing ISO transmission preserved approximately 75% and 92% of BOLD power in PING⁺ and PING⁻, respectively, but removed observed FC–SC alignment and weakened empirical correspondence. All in all, our results show that PING⁺ generates ISO through intrinsic gamma beating and even-order sigmoid mixing, whereas PING⁻ receives ISO through external noise; in both regimes, ISO transmission shapes BOLD FC.

Authors

Institutions

Publication Details

Journal
Zenodo (CERN European Organization for Nuclear Research)
Published
2026-11-11
DOI
https://doi.org/10.5281/zenodo.24141602
Primary Topic
Functional Brain Connectivity Studies
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

Infraslow generation and transmission shape BOLD connectivity in neural mass models

Borja Mercadal, Giulio Ruffini, Francesca Castaldo, Francisco Páscoa dos Santos
Zenodo (CERN European Organization for Nuclear Research)
Functional Brain Connectivity Studies
preprint

Infraslow generation and transmission shape BOLD connectivity in neural mass models

Borja Mercadal, Giulio Ruffini, Francesca Castaldo, Francisco Páscoa dos Santos
preprint en

Abstract

Resting-state functional connectivity (FC) is measured from slow blood-oxygen-level-dependent (BOLD) fluctuations, but their origin and its consequences for FC remain unresolved. We compared two sources of infraslow activity (ISO; 0.01–0.1 Hz) in pyramidal–interneuron gamma (PING)-like neural-mass networks: autonomous gamma interactions in noise-free networks (PING⁺) and regionally independent external noise in locally damped networks (PING⁻). Group-average Human Connectome Project diffusion-MRI structural connectivity (SC) and resting-state FC, distributed through the ENIGMA Toolbox, supplied the 82-region network and empirical benchmark. Frequency-selective interventions, calibrated Stuart–Landau (SL) comparators and BOLD decomposition separated generation from transmission. In PING⁺, carrier interventions and temporal prediction supported ISO generation by gamma beating. Converting beating between nearby gamma rhythms into a slow rate component requires even-order mixing, supplied here by the sigmoid and omitted from the first-order SL comparison. This endogenous source is constrained by SC, circuit parameters and tonic inputs; the coupled-equilibrium linearization contained no slow modes. In PING⁻, noise prescribes the ISO input and neural ISO persists without gamma exchange. Both regimes produced SC-aligned FC, but their FC patterns differed, and PING⁺ showed stronger empirical correspondence at the tested operating points and record lengths. With first-order transmitted rates and readout, noise-free SL⁺ produced negligible ISO, whereas noise-driven SL⁻ retained ISO and structural alignment. Suppressing ISO transmission preserved approximately 75% and 92% of BOLD power in PING⁺ and PING⁻, respectively, but removed observed FC–SC alignment and weakened empirical correspondence. All in all, our results show that PING⁺ generates ISO through intrinsic gamma beating and even-order sigmoid mixing, whereas PING⁻ receives ISO through external noise; in both regimes, ISO transmission shapes BOLD FC.

Zenodo (CERN European Organization for Nuclear Research)
Neuroelectrics Barcelona SLU (Spain) (ES)
Functional Brain Connectivity Studies
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon