While we sleep, our brains are busy organizing fresh memories into long-term storage—or at least, that’s the theory. Intriguing new research is bolstering this assertion, with evidence that our brains replay the day’s experiences during sleep, in what is an integral part of the memory-storage process.
In our brains, the conversion of short-term memories to long-term memories happens while we sleep, including during short naps. Neuroscientists refer to this process as “memory consolidation,” and it has been the subject of many studies, both on humans and animals, for decades.
There are plenty of moving parts to this hypothesized process, but the two key brain areas involved are the hippocampus and the neocortex. The hippocampus is responsible for learning and memory, and it’s a super-plastic part of the brain as its synaptic strengths can change quickly. The neocortex, on the other hand, is far less pliable, making it a reliable place to store long-term memories. Very simply, memory consolidation is a process in which the hippocampus delivers newly formed memories to the slower-learning neocortex, which is less susceptible to data loss, so to speak.
“It does so by quickly strengthening the connections in the pathways through the hippocampus that happen to interconnect those neocortical areas,” Beata Jarosiewicz, a senior research scientist at NeuroPace and a co-author of the new study, told Gizmodo. “Later, the hippocampus can then reactivate the neocortical neurons that were involved in processing all the components of those original events... and their reactivation is experienced as a memory.”
During sleep or rest, the repeated activation or “offline replay” of recent experiences is believed to consolidate these memories. In other words, “to make them stronger and to gradually incorporate them synergistically into our existing knowledge base in the neocortex, without disrupting the information that’s already there,” said Jarosiewicz, who previously worked on the research initiative BrainGate.
The new paper, published today in Cell Reports, adds to this discussion by presenting evidence of offline replay in the human brain during sleep, in a finding that’s strengthening the standard memory consolidation hypothesis. This process has been documented before in nonhuman animals, but the new study is the first to show it in human brains.
“Memory replay has been extensively described in non-human animals,” said Jarosiewicz, with neuroscientists able to track brain activity down to individual neurons. In humans, indirect evidence of this was seen using non-invasive tools like EEGs and FMRIs, she said, but the “recording technologies in those studies did not have the spatial resolution to directly test for replay at the level of neural firing rate patterns, as we have done here.”
For the new study, two participants with quadriplegia had intracortical microelectrode arrays implanted in their motor cortex, the part of the brain responsible for movement. This was done as part of a BrainGate pilot clinical trial, in which computer cursors, robotic prosthetic limbs, and other assistive devices can be controlled by a person’s thoughts. BrainGate is an academic research initiative involving Brown University, Massachusetts General Hospital, Case Western Reserve University, Stanford University, and the Providence VA Medical Center.
Both participants were asked to play a game similar to Simon, the classic memory game from the 1980s. Basically, the game involves four colored panels, which flash an increasingly complex pattern that participants must memorize on the fly. But instead of mimicking these patterns with their fingers, the participants were able to do so with a mind-controlled cursor, thanks to the implant. Both participants performed many repetitions of a standard sequence. The standard sequence was interspersed with outlier sequences, which served as the controls and only appeared twice each. The researchers monitored the two participants’ neural patterns while they played the memory game.
After a gaming session, the participants were asked to take a nap or simply rest. This gaming-sleep pattern was repeated several times, and the researchers recorded spiking activity in their neuronal clusters throughout the experiment.
Analysis of the data showed that the specific neuronal firing patterns seen during the gaming sessions matched those during sleep and wakeful resting. In a way, it was like the participants were unconsciously playing the game. The brain activity patterns recorded during rest more closely matched those seen while they carried out the learned color sequence compared to the outlier control sequences.
The researchers say this is the first direct evidence of learning-associated replay in the human brain. But plenty of work remains, as this study does not present evidence showing a causal relationship between offline replay and long-term memory consolidation.
“One thing this study didn’t address is the degree to which the strength of replay relates to the strength of learning,” Jarosiewicz told Gizmodo. “Future studies could probe this relationship further by, for example, re-testing participants on the repeated sequences again after their post-game nap. We would also like to look at how replay evolves over longer periods and different stages of sleep, including overnight sleep, and maybe even explore how replay might relate to phenomena like dreaming.”
Interestingly, the new research has implications for students or anyone else preparing for a memory-related activity. Pulling all-nighters and avoiding sleep prior to an exam, presentation, or interview is not advised, if this interpretation is correct. Instead, get a good night sleep, and consolidate those newly acquired memories!