Harmonizing brain rhythms: cortex-wide neuronal dynamics underpin quasi-periodic patterns in resting-state fMRI

Abstract Functional magnetic resonance imaging (fMRI) captures whole-brain activity fluctuations non-invasively in humans and animals. Beyond task/stimuli-locked responses, fMRI measures large-scale patterned activity during rest. An established method for identifying patterned activity in fMRI data, termed quasi-periodic pattern (QPP) analysis, identifies waves of activity which unfold over seconds and have consistent spatiotemporal characteristics. Notably, certain fMRI-QPPs are well-preserved across species and altered in various neuropsychiatric and neurodegenerative diseases. Yet, our collective understanding of their neural underpinnings is limited given the indirect nature of blood-oxygen-level dependent (BOLD) fMRI signals. Simultaneous measures of local field potentials have provided some affirmation that fMRI-QPPs have neural origins, but these point-measurements are limited to a handful of sites. Here, we use a unique multimodal implementation of simultaneous wide-field calcium (WF-Ca 2+ ) imaging and fMRI to investigate the neural origins of fMRI-QPPs.. We uncover a robust time-locked correlation between QPPs detected by cortex-wide fluorescent WF-Ca 2+ imaging of neural activity and QPPs of BOLD-fMRI. These data validate the hypothesis that BOLD QPPs derive from preceding slow waves of neural activity with regional and temporal precision.