During sleep, one brain region teaches another, transforming new information into long-lasting memories.
Credit: Pixabay/CC0 Public Domain |
What role do the stages of sleep play in memory formation? "We've known for a long time that useful learning occurs during sleep," says Anna Schapiro, a neuroscientist at the University of Pennsylvania. "You encode new experiences while awake, go to sleep, and when you wake up, your memory has been transformed in some way."
However, how new experiences are processed during sleep remains largely unknown. Schapiro, Penn Ph.D. student Dhairyya Singh, and Princeton University's Kenneth Norman have gained new insight into the process by developing a neural network computational model.
According to research published in the Proceedings of the National Academy of Sciences, the hippocampus teaches the neocortex what it learned as the brain cycles through slow-wave and rapid-eye movement (REM) sleep about five times per night, transforming novel, fleeting information into enduring memory.
"This isn't just a model of learning in the brain's local circuits. It is how one brain region can teach another brain region while sleeping, when there is no external guidance "says Schapiro, an assistant professor of psychology at Penn. "It's also a proposal for how we can learn gracefully as our environment changes over time."
Schapiro's research broadly focuses on human learning and memory, specifically how people acquire and consolidate new information. She's long suspected that sleep played a role in this, which she and her colleagues have been testing in a lab by recording what happens in the brain while participants sleep.
In addition, her team creates neural network models to simulate learning and memory functions. For this study, Schapiro and colleagues created a neural network model comprised of a hippocampus, the brain's center for new memories, tasked with learning the world's day-to-day, episodic information, and the neocortex, which is in charge of aspects such as language, higher-level cognition, and more permanent memory storage. The researchers can watch and record which simulated neurons fire when in these two areas during simulated sleep, then analyze those activity patterns.
The researchers used a brain-inspired learning algorithm they created to run several sleep simulations. The simulations revealed that during slow-wave sleep, the brain mostly revisits recent incidents and data, guided by the hippocampus, whereas during REM sleep, it mostly reruns previous events, guided by memory storage in the neocortical regions.
"During non-REM sleep, when the two brain regions connect, that's when the hippocampus is actually teaching the neocortex," says Singh, a second-year doctoral student in Schapiro's lab. "The neocortex then reactivates and can replay what it already knows during the REM phase," solidifying the data's hold in long-term memory.
He also believes that alternating between the two stages of sleep is important. "When the neocortex does not have the opportunity to replay its own information, the information there is overwritten. We believe that strong memory formation requires alternating REM and non-REM sleep."
Although some aspects of the model are still theoretical, the findings are consistent with what is known in the field. "We still need to test this," says Schapiro. "One of our next steps will be to conduct experiments to determine whether REM sleep truly brings up old memories and what implications this may have for integrating new information into your existing knowledge."
Because the current simulations are based on a typical adult getting a good night's sleep, they may not apply to other types of adults or those with less-than-ideal sleep habits. They also provide no insight into what is going on with children, who require different amounts and types of sleep than adults. Schapiro believes her model has great potential to answer some of these outstanding questions. "Having a tool like this allows you to go in many directions," she says, "especially because sleep architecture changes across the lifespan and in different disorders, and we can simulate these changes in the model."
In the long run, a better understanding of the role of sleep stages in memory could help inform treatments for psychiatric and neurological disorders characterized by sleep deficits. According to Singh, there may be implications for deep learning and artificial intelligence. "Our biologically inspired algorithm may pave the way for more powerful offline memory processing in AI systems," he says. This proof-of-concept study linking sleep and memory formation brings the field one step closer to these objectives.
What role do the stages of sleep play in memory formation? "We've known for a long time that useful learning occurs during sleep," says Anna Schapiro, a neuroscientist at the University of Pennsylvania. "You encode new experiences while awake, go to sleep, and when you wake up, your memory has been transformed in some way."
However, how new experiences are processed during sleep remains largely unknown. Schapiro, Penn Ph.D. student Dhairyya Singh, and Princeton University's Kenneth Norman have gained new insight into the process by developing a neural network computational model.
According to research published in the Proceedings of the National Academy of Sciences, the hippocampus teaches the neocortex what it learned as the brain cycles through slow-wave and rapid-eye movement (REM) sleep about five times per night, transforming novel, fleeting information into enduring memory.
"This isn't just a model of learning in the brain's local circuits. It is how one brain region can teach another brain region while sleeping, when there is no external guidance "says Schapiro, an assistant professor of psychology at Penn. "It's also a proposal for how we can learn gracefully as our environment changes over time."
Schapiro's research broadly focuses on human learning and memory, specifically how people acquire and consolidate new information. She's long suspected that sleep played a role in this, which she and her colleagues have been testing in a lab by recording what happens in the brain while participants sleep.
In addition, her team creates neural network models to simulate learning and memory functions. For this study, Schapiro and colleagues created a neural network model comprised of a hippocampus, the brain's center for new memories, tasked with learning the world's day-to-day, episodic information, and the neocortex, which is in charge of aspects such as language, higher-level cognition, and more permanent memory storage. The researchers can watch and record which simulated neurons fire when in these two areas during simulated sleep, then analyze those activity patterns.
The researchers used a brain-inspired learning algorithm they created to run several sleep simulations. The simulations revealed that during slow-wave sleep, the brain mostly revisits recent incidents and data, guided by the hippocampus, whereas during REM sleep, it mostly reruns previous events, guided by memory storage in the neocortical regions.
"During non-REM sleep, when the two brain regions connect, that's when the hippocampus is actually teaching the neocortex," says Singh, a second-year doctoral student in Schapiro's lab. "The neocortex then reactivates and can replay what it already knows during the REM phase," solidifying the data's hold in long-term memory.
He also believes that alternating between the two stages of sleep is important. "When the neocortex does not have the opportunity to replay its own information, the information there is overwritten. We believe that strong memory formation requires alternating REM and non-REM sleep."
Although some aspects of the model are still theoretical, the findings are consistent with what is known in the field. "We still need to test this," says Schapiro. "One of our next steps will be to conduct experiments to determine whether REM sleep truly brings up old memories and what implications this may have for integrating new information into your existing knowledge."
Because the current simulations are based on a typical adult getting a good night's sleep, they may not apply to other types of adults or those with less-than-ideal sleep habits. They also provide no insight into what is going on with children, who require different amounts and types of sleep than adults. Schapiro believes her model has great potential to answer some of these outstanding questions. "Having a tool like this allows you to go in many directions," she says, "especially because sleep architecture changes across the lifespan and in different disorders, and we can simulate these changes in the model."
In the long run, a better understanding of the role of sleep stages in memory could help inform treatments for psychiatric and neurological disorders characterized by sleep deficits. According to Singh, there may be implications for deep learning and artificial intelligence. "Our biologically inspired algorithm may pave the way for more powerful offline memory processing in AI systems," he says. This proof-of-concept study linking sleep and memory formation brings the field one step closer to these objectives.
Journal information: Proceedings of the National Academy of Sciences