Lisa Meyer-Baese ^1,2 Nmachi Anumba ^1,2 Taylor Bolt ^1,2 L Daley ^1,2 T J. Lagrow ² Xiaodi Zhang ^1,2 Nan Xu ^1,2 Wen-Ju Pan ^1,2 E H. Schumacher ^1,2 Shella Keilholz ^1*

• ¹ Biomedical Engineering, Emory University, Atlanta, GA, United States
• ² Georgia Institute of Technology, Atlanta, Georgia, United States

A few large-scale spatiotemporal patterns of brain activity (quasiperiodic patterns or QPPs) account for most of the spatial structure observed in resting state functional magnetic resonance imaging (rs-fMRI). The QPPs capture well-known features such as the evolution of the global signal and the alternating dominance of the default mode and task positive networks. These widespread patterns of activity have plausible ties to neuromodulatory input that mediates changes in nonlocalized processes, including arousal and attention. To determine whether QPPs exhibit variations across brain conditions, the relative magnitude and distribution of the three strongest QPPs were examined in two scenarios. First, in data from the Human Connectome Project, the relative incidence and magnitude of the QPPs was examined over the course of the scan, under the hypothesis that increasing drowsiness would shift the expression of the QPPs over time. Second, using rs-fMRI in rats obtained with a novel approach that minimizes noise, the relative incidence and magnitude of the QPPs was examined under three different anesthetic conditions expected to create distinct types of brain activity. The results indicate that both the distribution of QPPs and their magnitude changes with brain state, evidence of the sensitivity of these large-scale patterns to widespread changes linked to alterations in brain conditions.