Whole-brain normative modeling across multiple functional connectivity metrics in MEG
- 1Aarhus University
Abstract
Normative modeling has recently emerged in structural and functional neuroimaging as a promising approach to characterize neuropathology at the individual level. Trained on large healthy cohorts, these models predict expected measurements from covariates such as age and sex while quantifying normal population variability. Individual measurements can then be compared against the model to identify deviations associated with pathology. While most normative models of brain function have been based on functional MRI, EEG and MEG offer more direct measurements of brain activity at higher temporal resolution. Despite the promise, how to adequately perform normative modelling at the whole-brain level on these modalities remains challenging. We here propose M/EEG-FUNCOIN to identify age- and sex-related dimensions of change in functional connectivity from a plethora of FC measures from electrophysiology: amplitude envelope correlation, phase-locking value, and time-delay embedding covariance. Using resting-state MEG data from the Cambridge Centre for Aging and Neuroscience (Cam-CAN) dataset, we show that M/EEG-FUNCOIN can: (i) characterize age- and sex-related whole-brain functional connectivity patterns in greater detail than existing methods; and (ii) effectively capture meaningful deviations in connectivity in individuals that rank high and low in cognitive ability. Altogether, these results show great potential for applications in complex, network-level brain pathology.
Related
- Normative Modeling of Molecular-Enriched Functional Connectivity for Detecting Deviations from Healthy Brain Aging
- Parsing altered brain connectivity in neurodevelopmental disorders by integrating graph-based normative modeling and deep generative networks
- Marine Functional Connectivity Contributes to Ecosystem Functioning Across Multiple Scales
- Brain age prediction from structural and functional connectivity across the lifespan using connectome-based predictive modeling
- Hidden Markov Modeling Reveals Altered Functional Network Dynamics in Multiple Sclerosis
BibTeX
@misc{kobbersmed2026wholebrain,
title = {{Whole-brain normative modeling across multiple functional connectivity metrics in MEG}},
author = {J. R. L. Kobbersmed and C. Gohil and M. W. Woolrich and D. Vidaurre},
howpublished = {bioRxiv},
year = {2026},
month = oct,
doi = {10.64898/2026.10.21.760411},
url = {https://www.biorxiv.org/content/10.64898/2026.10.21.760411v1},
note = {CC BY},
}
RIS
TY - UNPB
TI - Whole-brain normative modeling across multiple functional connectivity metrics in MEG
AU - J. R. L. Kobbersmed
AU - C. Gohil
AU - M. W. Woolrich
AU - D. Vidaurre
PY - 2026
DA - 2026/10/21
PB - bioRxiv
DO - 10.64898/2026.10.21.760411
UR - https://www.biorxiv.org/content/10.64898/2026.10.21.760411v1
AB - Normative modeling has recently emerged in structural and functional neuroimaging as a promising approach to characterize neuropathology at the individual level. Trained on large healthy cohorts, these models predict expected measurements from covariates such as age and sex while quantifying normal population variability. Individual measurements can then be compared against the model to identify deviations associated with pathology. While most normative models of brain function have been based on functional MRI, EEG and MEG offer more direct measurements of brain activity at higher temporal resolution. Despite the promise, how to adequately perform normative modelling at the whole-brain level on these modalities remains challenging. We here propose M/EEG-FUNCOIN to identify age- and sex-related dimensions of change in functional connectivity from a plethora of FC measures from electrophysiology: amplitude envelope correlation, phase-locking value, and time-delay embedding covariance. Using resting-state MEG data from the Cambridge Centre for Aging and Neuroscience (Cam-CAN) dataset, we show that M/EEG-FUNCOIN can: (i) characterize age- and sex-related whole-brain functional connectivity patterns in greater detail than existing methods; and (ii) effectively capture meaningful deviations in connectivity in individuals that rank high and low in cognitive ability. Altogether, these results show great potential for applications in complex, network-level brain pathology.
ER -