Functional Neuroimaging in Sleep and Sleep Disorders
Functional neuroimaging characterizes the human brain function in time and space, using various techniques such as electroencephalography (EEG), positron emission tomography (PET), single photon emission computed tomography (SPECT), functional magnetic resonance imaging (fMRI) or near infrared spectroscopy (NIRS). In this chapter, we review the studies of human sleep using PET, SPECT or fMRI, all techniques which allow us to investigate the whole cerebral volume. In the following sections, we consider successively studies performed in normal subjects and in several sleep disorders.
The regional organization of brain activity is completely different in sleep than in wakefulness and reflects cellular processes similar to the ones described in sleeping animals.
Functional neuroimaging by PET has recently yielded original data on the functional neuroanatomy of human sleep. A more comprehensive discussion of these results, which is beyond the scope of the present chapter, can be found in a recent review paper (Maquet, 2000). These recent data describe the very reproducible functional neuroanatomy in sleep. The core characteristics of this 'canonical' sleep may be summarized as follows. In slow-wave sleep (SWS), the most deactivated areas are located in the dorsal pons and mesencephalon, cerebellum, thalami, basal ganglia, basal forebrain/hypothalamus, prefrontal cortex, anterior cingulate cortex, precuneus and the mesial aspect of the temporal lobe. Briefly, these findings are in keeping with the generation of non-rapid eye movement (REM) sleep in mammals, whereby the decreased firing in brainstem structures causes hyperpolarization of the thalamic neurons and triggers a cascade of processes responsible for the generation of various non-REM sleep rhythms (spindles, theta, and slow rhythm; see Chapter XXIV-7). The human data showed for the first time that the pattern of cortical deactivation was not homogeneous but predominated in various associative cortices, particularly the dorso-lateral and orbital prefrontal cortex. There seems to be a functional link between these cortical regions and sleep processes. Indeed, these areas are known to be involved in mood regulation and in various cognitive functions (such as planning or probability matching) that help to adapt individual behaviour and are known to deteriorate after a short deprivation of sleep (Harrison and Horne, 1998, 1999; Horne, 1988, 1993; Pilcher and Huffcutt, 1996).
During REM sleep, significant activations were found in the pontine tegmentum, thalamic nuclei, limbic areas:amygdaloid complexes (Maquet et al., 1996; Nofzinger et al., 1997), hippocampal formation (Braun et al., 1997; Nofzinger et al., 1997), and anterior cingulate cortex (Braun et al., 1997; Maquet et al, 1996; Nofzinger et al., 1997). The posterior cortices in temporo-occipital areas were also activated (Braun et al., 1997) and their functional interactions are different in REM sleep than in wakefulness (Braun et al., 1998). In contrast, the dorso-lateral prefrontal cortex and parietal cortex, as well as the posterior cingulate cortex and precuneus, were the least active brain regions (Braun et al., 1997; Maquet et al., 1996). Once again, these regional distributions of brain activity are in good accordance with the knowledge already acquired on sleep in animals (see Chapter XXIV-7). REM sleep is generated by neuronal populations of the meso-pontine reticular formation that monosynaptically activate the thalamic nuclei and in turn the cortex. Although early animals studies had mentioned the high limbic activity during REM sleep (Lydic et al., 1991), functional neuroimaging in humans highlighted the contrast between this activation of limbic, paralimbic and posterior cortical areas, and the relative quiescence of associative frontal and parietal cortices. This particular pattern of activation has been thought to account for the main characteristics of human dreaming activity (Hobson et al., 1998; Maquet, 2000; Maquet et al, 1996). The perceptual aspects of dreams may be related to the activation of the posterior (occipital and temporal) cortices and the emotional features to the activation of the amygdalar complexes and related mesio-temporal areas. In contrast, the lack of insight, the loss of time perception, and the amnesia on awakening may be related to the relative hypoactivation of the prefrontal cortex.
In the adult brain, it is believed that sleep periods participate in memory trace consolidation (Hennevin et al., 1995; Smith, 1995). If this were the case, the regional brain activity in sleep would not be fixed and stereotyped but would depend on previous waking experience. It was recently shown that waking experience indeed influences regional brain activity during subsequent sleep (Maquet et al., 2000). During REM sleep, several brain areas, activated during the execution of a serial reaction time (SRT) task during wakefulness, are significantly more active in subjects previously trained on the task than in non-trained subjects. The functional connectivity of these brain regions was also examined (Laureys et al., unpublished data). One of the reactivated areas, the left premotor cortex, was functionally more tightly correlated with the posterior parietal cortex and the supplementary motor area (SMA) during post-training REM sleep than in 'typical' REM sleep. It was hypothesized that this increase of functional connectivity during post-training REM sleep reflects some processing of the memory traces recendy acquired during wakefulness and embodied in the parietal-premotor-SMA network. It remains to be shown that these processes lead to memory trace consolidation.