In groundbreaking study, researchers publish brain map showing how decisions are made

Each coloured dot in this image represents a neuron that is firing during the decision-making process. (Daniel Birman/International Brain Laboratory via CNN Newsource)

Neuroscientists from 22 labs joined forces in an unprecedented international partnership to produce a landmark achievement: a neural map that shows activity across the entire brain during decision-making.

The data, gathered from 139 mice, encompass activity from more than 600,000 neurons in 279 areas of the brain — about 95% of the brain in a mouse. This map is the first to provide a complete picture of what happens across the brain as a decision is made.

“They have created the largest dataset anyone has ever imagined at this scale,” said Dr. Paul W. Glimcher, chair of the department of neuroscience and physiology and director of the Neuroscience Institute at New York University’s Grossman School of Medicine, of the researchers.

In the field of neuroscience, “this is going to go down in history as a major event,” Glimcher, who was not involved in the new research, told CNN.

To construct the map, researchers first created a standardized procedure to be shared across laboratories and then tracked neural activity in mice as the rodents responded to visual prompts, integrating all the data gathered by each lab. Seven years in the making and presented in two studies, the findings were published on September 3 in the journal Nature.

“There are basically two big results, which is why we have two papers,” said Alexandre Pouget, a full professor in basic neuroscience at the University of Geneva. One study outlined the widespread distribution of electrical activity related to decision-making. The other used the data to evaluate how expectations shape choices. Pouget is a coauthor of the first study and senior author of the second.

“We started from scratch,” he told CNN. “Nobody had ever attempted to do something like this before.”

Prior research suggested that small clusters of neurons fire in only some parts of the brain during decision-making, mostly in areas related to sensory input and cognition. But the new map reveals that neural activity is far more widespread, with electrical signals pinging across nearly all of the mouse’s brain during different stages of decision-making.

From a single neuron to thousands at once

For decades, scientists have studied brain activity during certain tasks by using electrodes that record electrical pulses from single neurons. But recording one neuron at a time is difficult and slow; several months of work would yield results from around 100 neurons, making the technique best suited for studying highly targeted regions of the brain.

Over the past decade, neuroscience took a giant leap forward with the development of digital neural probes called Neuropixels, which can monitor thousands of neurons at once. These sensitive electrodes were an essential tool for creating the new map.

“We went from looking at just a few hundred neurons in one area to 600,000 neurons in all brain regions,” Pouget said.

In the experiments, mice wore electrode helmets while turning a tiny steering wheel to control the movement of a black-and-white striped circle on a screen. The circle briefly appeared on either the left side or the right side of a screen, and mice that successfully steered the circle to the center received a reward of sugar water. As the mice responded to what they saw, Neuropixels probes recorded electrical signals in their brains.

According to the map, activity first spiked toward the back of the brain, in areas that process visual input. Activity then spread across the brain, with motor-controlling areas lighting up as a mouse’s decision culminated in movement. Widespread brain activity followed when the mouse got its sugary reward.

“It’s not just a few areas involved in this, but a very large network of areas that work together,” Pouget said. Knowing how much of the brain is involved in decisions will help researchers conduct more targeted studies of complex behavior, the study authors reported.

Researchers also included an extra challenge for the mice. Sometimes the circle was faint, or nearly invisible. To decide which way to turn the wheel for a reward, a mouse would have to recall what it saw during earlier attempts.

“That’s called prior knowledge,” Pouget said. “Every decision you make is made that way.”

Neuroscientists previously hypothesized that the brain accesses prior knowledge early in decision-making, “so that as soon as you start processing your sensory stimulus, you do it in the context of what you expect,” Pouget said.

The brain map demonstrated that this prediction was correct, he noted.

‘A Sloan Digital Sky Survey for the brain’

Just as large-scale international collaborations in other disciplines have transformed how science is conducted, the scope of the work that produced the brain-activity map is a game changer for neuroscience, Glimcher said.

“Traditionally, the biological sciences have been a lab-by-lab science,” unlike the multi-lab collaborations that often occur in physics and astronomy, he said. A notable example is the Sloane Digital Sky Survey, which involves hundreds of astrophysicists and astronomers and has produced the most detailed 3D maps of the universe ever made, encompassing over one-third of the night sky.

“The Sloan Digital Sky Survey revolutionized the way we gather astronomical data and distribute it amongst astrophysicists,” Glimcher said. The vision of the research organization behind the brain map — the International Brain Laboratory — “was to build a Sloan Digital Sky Survey for the brain.”

Ideally, added Pouget, a cofounder of the IBL, this map will be the first of many large-scale collaborations between neuroscientists: “We’re really hoping that this is going to inspire other groups to start working with this kind of approach.”