Can physiological signals estimation improve fMRI connectivity assessment in anesthetized rats?
Résumé
Resting-state functional magnetic resonance imaging (fMRI) is a common translational method to evaluate brain functional connectivity (FC) in preclinical neurologic applications. However, unlike human studies, preclinical fMRI studies are usually performed under anesthetic condition. The confounding effect of the anesthesia on fMRI measurements is not clear. Indeed, a controversy exists between those who claim that anesthesia change FC (Paasonen et al., 2018) and other who maintain that FC is preserved (Becq et al., 2020). Our hypothesis is that physiological changes induced by anesthesia could explain this discrepancy. While in human studies, fMRI analysis generally includes cleaning of physiological-noise (Glover et al., 2000; Kassinopoulos & Mitsis, 2019), preclinical study processing is rather often limited to a low-pass temporal filtering. We aimed first to demonstrate that the global MRI signal of animal brain includes, like in human, a physiological component. Second, that a normalization of this component based on global signal can reduce variability of FC assessment.
| Origine | Fichiers produits par l'(les) auteur(s) |
|---|---|
| Licence |