Journal Club

Highlighting recent, timely papers selected by Academy member labs

Several brain regions help us anticipate what’s going to happen next

Vertical slices of the brain, imaged at different locations, reveal a timescale gradient for anticipation. Timescales are short at the back of the brain (cool blues) and longer at the front (warm reds). Image credit: Christopher Baldassano

Vertical slices of the brain, imaged at different locations, reveal a timescale gradient for anticipation. Timescales are short at the back of the brain (cool blues) and longer at the front (warm reds). Image credit: Christopher Baldassano

If you watch a movie clip on repeat, many areas of your brain begin to anticipate upcoming events onscreen, according to a recent study published in eLife. The first time you see the clip, your brain forms a distinctive neural pattern. Then, as you become familiar with the clip on subsequent viewings, that neural pattern shifts earlier, demonstrating some level of expectation as to what’s coming next. The findings speak to more than just how we react to repeat viewings of The Grand Budapest Hotel (the film used in this study). They could inform thinking about how the human brain is organized.

Researchers tracked brain activity in 30 participants, each watching the same 90-second film clip six times. Functional magnetic resonance imaging (fMRI) scans of their brains revealed activity patterns shifting earlier in repeat viewings of the clip. The scans also showed that the degree of anticipation varied by brain region. The frontal cortex, for example, could anticipate up to 15 seconds in advance, perhaps foreseeing movie plot changes, while back-brain cortical areas could only anticipate about 1 second ahead, likely predicting changes in words or simple images.

Together, these findings point to “an organized pattern in the brain,” says coauthor Christopher Baldassano, a cognitive neuroscientist at Columbia University in New York. “Parts near [the] front of the brain are showing longer anticipation of things farther in the future. So the brain is able to create these anticipatory signals at multiple timescales.”

The fMRI scans measured activity in thousands of small patches of the brain simultaneously. For each small patch, several millimeters in size, the coauthors watched neural activity rise and fall as a participant watched the clip. Researchers then used a machine learning model to identify the shared, average pattern of activity across all 30 participants when they first watched the clip. For example, brain activity might increase by 3 units, then decrease by 5, then increase again, by 2, over the duration of the clip, Baldassano explains.

When the participants watched the clip a second time, the researchers watched that same patch of brain and gauged whether average neural activity went up and down earlier than it did on the first viewing. The answer seems to be yes. “We found that lots of parts of the brain show this kind of anticipation,” Baldassano says. Even areas traditionally associated with sensory perception and other much simpler tasks anticipated to some degree, though the amount varied by region, ranging from a couple of seconds to about 15.

The finding is consistent with evidence from memory studies asking how much information about the past different brain regions can hold onto. That work found hierarchies moving from the front to the back of the brain, where frontal regions can keep track of tens of seconds, compared to only a few seconds at the back of the brain, Baldassano explains.

A hierarchical view of the brain is very different from the traditional, modular view, he notes. In the traditional view, there are systems that process raw sensory inputs, such as sights or sounds, and there are separate systems that call up memories or make plans. The latest findings blur the lines between those systems, showing that regions known for processing simple pieces of visual information, such as the visual cortex, can also anticipate what’s coming up soon, even if just by a few seconds.

“There is a lot of interest in how the brain makes predictions about the future over different scales and how this information is organized over the brain,” says neuroscientist Kimberly Stachenfeld at DeepMind in London, who was not involved in the study. This paper, Stachenfeld notes, starts to unpack that, by showing how different timescales of prediction are represented across the brain. She adds that these latest results are consistent with prior studies suggesting a gradient or hierarchy moving from the front to the back of the brain. But looking specifically at anticipation in this context is new, she says.

“These results open up a lot of exciting possibilities, such as looking at whether such prediction also generalizes to similar but not identical experiences,” says Iva Brunec, a postdoc in cognitive neuroscience at Temple University, in Philadelphia, PA. It will also be intriguing to investigate “whether people systematically differ in how far ahead into the future their brain patterns might reach,” she adds.

Among the next steps: exploring whether participants in studies like this one consciously know they are anticipating, and whether they use that anticipation to make plans. In this study, all 30 people passively watched the film clips without additional directions. But one future path of study, which coauthor Mariam Aly, also a cognitive neuroscientist at Columbia is spearheading now, will ask participants to generate predictions about the near and far future while completing a task. That should enable Aly and collaborators to see if people’s behavior and judgements are related to what their brain is anticipating, in the near or far future.

Other recent papers recommended by Journal Club panelists:

The metabolic network of the last bacterial common ancestor

Whitefly hijacks a plant detoxification gene that neutralizes plant toxins

Association with a novel protective microbe facilitates host adaptation to a stressful environment

Early or late gestational exposure to maternal immune activation alters neurodevelopmental trajectories in mice: an integrated neuroimaging, behavioural, and transcriptional study

New hominin remains and revised context from the earliest Homo erectus locality in East Turkana, Kenya

The interplay between top-down interventions and bottom-up self-organization shapes opportunities for transforming self-governance in small-scale fisheries.


Categories: Journal Club | Neuroscience | Psychological and Cognitive Sciences and tagged | | | | | | | | |
Print Email Comment

Leave a Comment

Your email address will not be published. Required fields are marked *