Predictability of arousal in mouse slow wave sleep by accelerometer data.

Area of Science:

• Neuroscience
• Sleep Science
• Biophysics

Background:

• Arousals, characterized by brief awakenings during sleep, are poorly understood, especially their underlying physiological mechanisms.
• Subtle body movements associated with arousals are often imperceptible and not typically included in sleep studies.
• Current methods for studying arousals often rely on human electroencephalography (EEG), which is not always suitable for animal models or non-invasive studies.

Purpose of the Study:

• To characterize and predict arousals during slow-wave sleep (SWS) in mice using accelerometer records (AR).
• To investigate the relationship between body movement patterns and neural activity preceding arousals.
• To explore the potential of recurrence quantification analysis (RQA) for understanding sleep dynamics.

Main Methods:

• Recorded accelerometer data (AR) and local field potentials (LFP) from the hippocampus (CA1 region) in mice during SWS.
• Utilized recurrence quantification analysis (RQA), specifically the determinism (DET) quantifier, to analyze AR and LFP time series.
• Identified SWS stages using hippocampal LFP signals and analyzed AR dynamics preceding arousals.

Main Results:

• A universal dynamic pattern in AR data, preceding arousals during SWS, was identified for the first time.
• Predictability of arousals using DET analysis on AR data achieved nearly 90% accuracy.
• Analysis of hippocampal LFP data yielded an 88% success rate in predicting arousals, highlighting the significance of movement patterns.

Conclusions:

• Accelerometer data reveals predictable dynamical patterns preceding sleep arousals, offering insights into the sleep-wake switch.
• The findings suggest a potential link between respiratory changes and neural states driving arousal events.
• AR analysis combined with other physiological data presents a promising non-invasive approach for studying sleep physiology and pathology.