In the last decades, science has taught us that the mammalian brain isn’t always entirely awake or asleep. Dolphins can swim with one hemisphere asleep while the other is alert, and some neurons in sleep-deprived rats can “switch off” while the animals are still awake. In humans, this so-called “local sleep,” in which specific neuronal populations take a nap while the rest of the brain is awake, has been more challenging to study, since the invasive methods used to track it in other mammals cannot be used on people.
A new study published July 21 in PNAS seems to have overcome this challenge. By simultaneously mapping human brain signals measured with two different methods (one with good temporal resolution and the other with good spatial resolution), the team pinpointed the waking or sleeping state of neuronal populations at the local level. The achievement made it possible to identify which brain regions are the first to fall asleep and which are the first to wake up, and experts say it promises to be a valuable tool for studying sleep in humans.
Traditionally, sleep in humans has been studied using electroencephalography (EEG), which measures the electrical activity of the brain via electrodes placed on the scalp. EEG is good at measuring rapid changes, but it has very poor spatial resolution, explains Chen Song, a brain researcher at Cardiff University who participated in the new study while a postdoc at the University of Wisconsin-Madison. Thus, although scientists have learned a lot about sleep signals from EEG, this technique tells us little about local sleep.
Song and her colleagues decided to pair EEG with a complementary technique, functional magnetic resonance imaging (fMRI), which measures blood flow in the brain as a proxy for neuronal activity. In contrast with EEG, fMRI is not good at measuring short, fast changes, but it can distinguish brain activity at a resolution of a cubic millimeter, explains Song. The team decided to explore if the neural signals typical of sleep (slow waves and bursts of oscillations known as sleep spindles) often found in EEG measurements could be mirrored by patterns provided by the fMRI.
See "Sleep’s Kernel"
Using data gathered for one of their previous studies, the team analyzed the brain activity of 36 people who fell slept wearing an EEG cap inside an fMRI scanner for one hour. By looking at the oscillations of the blood flow activity and comparing it with the EEG data, Song and her colleagues found that the typical electrical wave patterns of sleep were mirrored by the blood oxygen responses recorded by the fMRI.
The researchers further found that these blood flow oscillations had distinct spatiotemporal patterns across the brain, suggesting that some regions enter a sleeping state earlier than others as we drift off. For instance, in their analysis, the thalamus was the first region to show sleep-associated blood flow patterns, a finding consistent with previous work based on EEG data reporting that this region may shut down earlier than other regions during the transition to sleep.
They found a separate spatiotemporal pattern when measuring brain activity as the volunteers woke up. For example, the frontal cortical regions of the brain may be among the first to wake up. This is different from what researchers had previously thought based partially on animal research but mostly on theoretical reasoning, says Song, as these regions are associated with cognitive processing “and a lot of time when you wake up you [may] have the feeling you cannot [concentrate] at all.” However, Song acknowledges that sleeping inside the fMRI scan “is very unnatural,” and it is possible that people experience a lighter sleep, resulting in these observations.
See "Perchance to Dream"
This study offers “a new perspective” for studying the brain at the local level, says Vanderbilt University electrical engineer Catie Chang, who was not involved in this study, and the methods the authors used could help “put together a more complete picture” of what happens in the brain when we fall asleep.
Monitoring local sleep using fMRI could also be valuable in improving our understanding of sleep disorders. For instance, researchers can begin to ask what it looks like in healthy volunteers who have “no trouble falling asleep versus people who might have difficulty sleeping,” Chang says. And Song suggests the technique could be especially useful in understanding whether local sleep and local wakefulness have any function in humans, since currently “we don’t know anything about that.”
