CSF flow and neural activity associated with behavioral and physiological arousal state transitions

2567 

Abstract Submission 

Danlei Chen¹, Beverly Setzer², Laura Lewis³

¹MIT, Cambridge, MA, ²Boston University, Boston, MA, ³Massacusetts Institute of Technology, Cambridge, MA

Danlei Chen  
MIT
Cambridge, MA

Beverly Setzer  
Boston University
Boston, MA

Laura Lewis, Ph.D.  
Massacusetts Institute of Technology
Cambridge, MA

Cerebrospinal fluid (CSF) is an important component of the central nervous system that maintains essential biochemical and mechanical functions of the brain. Recent studies have found large waves of CSF flow in the fourth ventricle during non-rapid eye movement sleep (Fultz et al., 2019) and intense sensory stimulation (Williams et al., 2023), suggesting that CSF flow can be modulated by neural activity due to neurovascular coupling. In addition, autonomic arousal and global vascular dynamics can shape CSF flow (Picchioni et al., 2022). Here, we used accelerated, high-resolution fMRI to measure physiological and neural dynamics in the human brain during spontaneous transitions of both behavioral arousal and physiological arousal states. We hypothesized that arousal state transitions are associated with large-scale changes in CSF flow and whole-brain neural activity (Lewis et al., 2015; Raut et al., 2020; Setzer et al., 2022). We further used the high temporal resolution of these data to explore the temporal sequence across whole-brain networks that drive physiological changes during physiological arousal state transitions.

Ten participants who restricted their sleep to four hours the night prior to the experiment performed a behavioral task with fast fMRI at 7-Tesla (TR = 247 ms; 2.5mm isotropic voxels). Preprocessing included slice-time and motion correction. During the scan, participants were instructed to keep their eyes closed and press a button for every breath in and breath out. Behavioral arousals were defined as the first response after at least 20 s of unresponsiveness (Fig. 1B). Physiological arousals were defined as the increase in respiratory volume. Any arousal events that included excessive motion (>0.3 mm) were excluded from the analysis. Using fast fMRI to measure flow-related enhancement, we extracted CSF flow signals in the fourth ventricle by selecting the brightest voxels in bottom edge slices (Fig. 1A).

As hypothesized, we observed large-scale changes in CSF flow during the transitions of behavioral arousal states (Fig. 1C). Compared to the control region, CSF flow showed a significant increase in the 3.1 to 12.3 s window; i.e. the onset of arousal-locked increase in CSF flow appeared after the onset of arousal. The transition of behavioral arousal states also elicited an increase in respiratory volume (Fig. 1D). These results showed a system-wide changes in physiology and neural activity during behavioral arousal transitions. Next, we investigated transitions of arousal states defined by increases in respiratory volume. As hypothesized, we observed large changes in cortical, subcortical, and CSF signals during the transition of physiological arousal states (Fig. 2A). To characterize a temporal sequence of neural activity associated with changes in respiratory signals, we calculated cross-correlation to identify their maximal correlation coefficient and the corresponding temporal lag. We observed that most cortical regions were negatively correlated with and followed changes in respiratory volume, while subcortical regions were positively correlated with and preceded changes in respiratory volume (Fig. 2B). This result suggests a possible temporal sequence such that neural activity of the subcortical regions might elicit physiological changes which subsequently drive changes in neural activity and vascular constriction of cortical regions.

Using accelerated fMRI, we identified CSF flow increases locked to behavioral and physiological arousal state transitions, which were also tightly coupled with both BOLD and respiratory signals. Further investigation also revealed a network-level temporal sequence which identified subcortical regions as a potential driver for physiological changes which in turn drives changes in cortical signals. Together, these results identify temporal sequences of neural and hemodynamic signals that couple CSF flow to behavioral and physiological arousal states.

Subcortical Structures ²

Sleep and Wakefulness ¹

Brainstem

Cerebral Blood Flow

Cortex

FUNCTIONAL MRI

Sub-Cortical

Thalamus