Summary: Every decision begins subtly, as the brain weighs options long before action. Researchers have now shown that, despite individual differences in neuron activity, a shared underlying structure guides the brain toward unified decisions.
By training macaques in a color-choice task and recording neural activity, scientists discovered that neuron responses are shaped by a common “potential landscape” that varies with task difficulty. This study offers a new model for how the brain organizes complex decisions and may inform future understanding of psychiatric conditions that disrupt decision-making.
Key Facts:
- Shared Structure: Diverse neurons follow a common potential landscape that shapes their collective decision-making.
- Difficulty Modulation: Easier decisions show steeper neural “slopes,” promoting quick choices; harder ones flatten the landscape.
- Clinical Implications: Understanding this coordination may shed light on disorders like schizophrenia and bipolar disorder.
Source: Princeton University
Every decision begins invisibly.
Long before someone acts, the brain is already hard at work gathering evidence, weighing options, and gradually committing to a choice. But even when faced with the same evidence, people can arrive at different outcomes, especially when the decision is difficult.
Two different drivers in rush hour traffic, for example, see the same congested road, yet one might speed up to merge while another cautiously brakes.
How the brain, made up of billions of specialized cells, makes these split-second decisions has largely been a mystery, though.
Now, new findings from Princeton University, in collaboration with researchers at Cold Spring Harbor Laboratory, Stanford University, and Boston University, shed light on how diverse brain cells come together to guide a unified decision.
The researchers found that while individual neurons have perplexingly complex responses, their activity is shaped by a shared structure that ultimately guides the brain toward a unified choice.
The findings were published in the journal Nature on June 25.
Classic experiments in neuroscience have shown that the brain maps simple sensory information, like basic shapes or sounds, in predictable ways. A black rectangle rotated at a 45-degree angle will activate a specific group of cells in the visual cortex.
Change the angle slightly, though, and a different group lights up. But decisions, especially when tied to action, are more complicated than distinguishing slightly different tones or shapes, making it difficult for researchers to identify the neural code that guides decision making.
To overcome this challenge, the research team trained rhesus macaques to determine which color (red or green) was more dominant on a checkered screen. Easy trials were clear-cut, but ambiguous ones required careful deliberation.
As the monkeys considered their choice, researchers recorded activity from nerve cells in the dorsal premotor cortex, a brain region involved in translating decisions into actions.
They found that neurons responded very differently, even within the same trial, suggesting a high degree of “heterogeneity” or variability, in the neural code for decisions.
“The widespread assumption is that this heterogeneity reflects the complex dynamics involved in cognition,” said Tatiana Engel, Ph.D., associate professor at the Princeton Neuroscience Institute and the senior author of the study.
“But surprisingly, we found that this apparent complexity arises from a very different neural coding principle.”
To explain this diversity, the team developed a flexible computational model that revealed two critical features driving each neuron’s behavior: 1) tuning: when and to what kind of decision a neuron tends to respond; and 2) neural dynamics: represented by a “potential landscape” that guides activity.
In this model, valleys in the landscape represent stable decision that’s been made. As neural activity unfolds, it’s like a ball rolling across the terrain: steeper slopes push activity more decisively toward a choice.
When fitted to real data, the model showed that tuning remained consistent across easy and hard trials, but the shape of the potential landscape changed. In easier tasks, the landscape was steep, leading to quicker, more confident decisions. In harder tasks, the terrain was flatter and more susceptible to noise, increasing the chances of mistakes.
Though each neuron had a different individual response, they all appeared to share the same underlying potential landscape.
“Think of it like a group of skiers descending a mountain,” Engel said.
“Each prefers a slightly different path, but all are shaped by the same slope beneath them. Similarly, each neuron has its own preference and activity, but the group of cells collectively in the premotor cortex takes a coordinated journey and gradually settles into a stable state that represents the decision.”
Understanding how neurons collaborate to make decisions could offer deeper insight into fundamental brain function, and how it goes awry in disorders such as schizophrenia or bipolar disorder, where decision-making processes are altered.
With a new model in hand, Engel and her colleagues now plan to explore how different types of neurons, and the ways they connect, contribute to the diverse tuning and distinct phases of decision-making they observed.
“Every decision is unique,” Engel said. “But by digging down to the level of single trials and single neurons, we can start to make sense of it.”
About this decision-making and neuroscience research news
Author: Daniel Vahaba
Source: Princeton University
Contact: Daniel Vahaba – Princeton University
Image: The image is credited to Neuroscience News
Original Research: Open access.
“The dynamics and geometry of choice in the premotor cortex” by Tatiana Engel et al. Nature
Abstract
The dynamics and geometry of choice in the premotor cortex
The brain represents sensory variables in the coordinated activity of neural populations, in which tuning curves of single neurons define the geometry of the population code.
Whether the same coding principle holds for dynamic cognitive variables remains unknown because internal cognitive processes unfold with a unique time course on single trials observed only in the irregular spiking of heterogeneous neural populations.
Here we show the existence of such a population code for the dynamics of choice formation in the primate premotor cortex.
We developed an approach to simultaneously infer population dynamics and tuning functions of single neurons to the population state.
Applied to spike data recorded during decision-making, our model revealed that populations of neurons encoded the same dynamic variable predicting choices, and heterogeneous firing rates resulted from the diverse tuning of single neurons to this decision variable.
The inferred dynamics indicated an attractor mechanism for decision computation.
Our results reveal a unifying geometric principle for neural encoding of sensory and dynamic cognitive variables.
