Driving through town, you’re getting ready for a right turn that’ll take you by your favorite coffeeshop. But then, to the left, you see a new sign: Fresh Pie. Do you continue on towards coffee, or do you yank the wheel left for an unexpected shot at a sweet treat?
The answer depends on timing — and on your mental commitment to that latte. Can you safely merge over soon enough to switch your destination, or are you already in the right lane with your blinker on? Have you been anticipating that coffee all morning, or was it more of a passing whim?
A new study in mice outlines the neural basis of this phenomenon. It helps explain how the brain can maintain its commitment to a decision in the face of distraction, but also remain open to new information that might change the optimal response, researchers report
April 19, 2021 in Nature Neuroscience. And it proposes a new model for flexible communication between sensory and motor areas of the brain.
“There are very few studies showing the causal mechanisms that explain decision making — and this is one of them,” says Lorenzo Fontolan, a postdoc in group leader Sandro Romani’s lab. Fontolan co-led the study alongside Arseny Finkelstein, a postdoc in the lab of senior group leader Karel Svoboda.
When you do anything in response to information about the world around you, messages travel from sensory areas of the brain to motor areas. But there’s a lag between when you choose an action and when you actually take it. And in that interval, distractions (like a sign for pie when you’d decided on coffee) can pop up.
Past studies had suggested that distracting signals might be suppressed in the sensory cortex before ever making their way to the motor cortex. That is, some part of your brain might note the new sign, but not tell the brain region that’s responsible for planning future actions anything about it. Instead, the new research suggests those signals do get through to the motor cortex, changing the pattern of neural firing in that part of the brain. It just doesn’t always lead to a corresponding change in behavior.
“We found that the motor cortex listens to the sensory cortex through the whole period,” says Finkelstein. “The motor cortex is constantly monitoring what the sensory cortex is telling it.”
The researchers trained the mice to lick to the left or the right to get a food reward, depending on whether the sensory cortex was being stimulated or not. Then, they added distractions —additional conflicting stimulations of the sensory cortex that were occurring later in time, but still before the mice had a chance to respond to the first stimulus.
The sooner that distracting information came into the brain, the more likely it was to switch the behavioral outcome, the team noted. Mice tended to ignore late-breaking distractions. Timing wasn’t everything, though. The strength of a mouse’s commitment to the first decision also made a difference. A solid decision to take a particular action showed up in the motor cortex as a strong, consistent pattern of neural firing — one that was harder to disrupt with new incoming signals.
To investigate the underlying neural mechanisms further, Fontolan built an artificial neural network that mimicked the activity of actual neurons in the motor cortex of the mouse brain. Data from the neural network suggested that in order to switch a behavior, the new sensory signals coming into the motor cortex must be strong enough to cross an imaginary barrier, Fontolan says.
That barrier builds up over time, and can vary in height depending on how entrenched the original planned action is in the motor cortex. If the mouse wasn’t particularly committed to the first action, for example, the barrier would be low — it would be relatively easy for conflicting input to change the outcome. If the barrier were higher, it would be harder. But a strong enough new signal from the sensory cortex could still swamp the barrier like water rising over a dam, switching the decision.
This study looked at a relatively straightforward binary choice, but in the future, the researchers hope to investigate how the brain handles more complex decisions. They also plan to study the learning processes that underlie decision-making, such as how the brain links a specific stimulus with an appropriate response in the first place.
Arseny Finkelstein, Lorenzo Fontolan, Michael N. Economo, Nuo Li, Sandro Romani and Karel Svoboda. “Attractor dynamics gate cortical information flow during decision-making,” Nature Neuroscience, April 19, 2021. Doi: 10.1038/s41593-021-00840-6.