A continuum model for the dynamics of the phase transition from slow-wave sleep to REM sleep

The cortical transition from the slow-wave pattern of sleep (SWS) to the rapideye-movement (REM) pattern is a dramatic feature of the somnogram. Indeed, the change in the electrocorticogram (ECoG) is so abrupt that the moment of transition usually can be identified with a time-resolution of about one second [8, 37]. Although the neuromodulatory environment and electroencephalographic patterns recorded during the steady states of SWS and REM have been well described [16, 30], the dynamics of the transition itself has been described only in a qualitative observational fashion [12], and has not been the focus of detailed quantitative modeling. In SWS, the rat cortex shows predominant activity in the delta (∼1–4 Hz) band. This pattern shifts to an intermediate sleep state (IS)—sometimes termed “preREM”—where the cortical activity shows features of both SWS and REM, lasting 10–30 seconds [7, 12, 25, 28]. This is followed by an abrupt transition to the REM state, characterized by strong theta (∼5–8 Hz) oscillation [2], and loss of delta power. The main effector of the cortical transition from SWS to REM is believed to be a linear progressive increase in cholinergic input into the cortex from the brainstem (mainly from the pedunculo-pontine tegmentum area), acting via the thalamus or basal forebrain [18, 34].