This repository accompanies “No increased brain age gap in persistent post-COVID-19 headache.” We trained a structural MRI brain age model on a large, multi-dataset healthy sample and applied it to adults with persistent post-COVID headache (n=53) and matched controls (n=48), with follow-up in patients (n=30). Morphometric features were extracted via FastSurfer and harmonised across scanners with ComBat-GAM; statistical analyses adjusted for age, sex, eTIV, and Euler number. Results showed no cross-sectional elevation in brain age gap and no systematic longitudinal change. As a robustness check, we replicated the analysis with the external deep-learning model Pyment, obtaining qualitatively similar null findings.
The Pyment brain age model weights and code are available from the repository: https://github.com/estenhl/pyment-public.
We included NeuroHarmonize (ComBat-GAM) in the repository, since we included in it harmonization assumptions checks (https://github.com/rpomponio/neuroHarmonize).
If you found this paper or repository useful, please reference:
[TO BE UPDATED]