An extensive brain map shows how regions collaborate during complex decision-making.
Imagine you’re driving on a foggy night, searching for a familiar turn-off with barely any visibility. You vaguely recall that the road usually bends left after a big oak tree. Even though you can’t clearly see the tree in the mist, you decide to turn left, trusting your memory. In that moment, your brain is frantically blending past experience with the faint visual cues in front of you to make a decision. From trivial choices like picking a coffee flavor to split-second life-or-death calls on the road, our decisions constantly draw on a tapestry of sensory information and prior knowledge. But what does that decision-making process look like inside the brain?
For the first time, scientists have managed to watch an entire brain make a decision in real time – not in a human, but in a mouse. In a landmark neuroscience experiment, an international team of researchers (including at UCLA) unveiled the first complete map of brain activity during decision-making, covering nearly the whole mouse brain . The project, led by the International Brain Laboratory (IBL), recorded the electrical activity of hundreds of thousands of neurons across the brain as mice engaged in a decision task . The resulting brain-wide map, published in Nature in September 2025, offers an unprecedented window into how different brain regions work together when an animal makes up its mind .
Mapping a Mouse’s Decision
The feat was made possible by a massive collaboration. The IBL was launched in 2017 as a worldwide coalition of neuroscientists with a bold goal: to link brain activity to behavior in ways no single lab could achieve . In the past, studies of decision-making had to pick off small bits of the problem – one lab might record neurons in the visual cortex, another in the frontal cortex, each using different tasks and methods. The IBL instead united 12 laboratories across the US and Europe to tackle one experiment together, using the same methods and sharing all their data. “The brain is the most complex structure we know of in the universe and understanding how it drives behavior requires international collaboration on a scale that matches that complexity,” said IBL co-founder Tom Mrsic-Flogel . Inspired by big science efforts like CERN’s particle physics teams and the Human Genome Project, the IBL built a standardized pipeline so that data collected in New York or London or Lisbon would all fit together seamlessly .
At the heart of the experiment was a clever behavioral task – essentially, a video game for mice designed to probe decision-making. Each mouse was placed in front of a little screen and a tiny steering wheel. A pattern of lines would briefly appear on either the left or right side of the screen, and the mouse had to turn the wheel to center the pattern in order to get a reward (usually a sip of water or juice) . If the mouse guessed wrong, it got a brief timeout or an unpleasant noise instead of a treat . As training progressed, the task gradually became harder: sometimes the striped pattern was made so faint that the mouse could barely see it at all, essentially forcing the animal to guess which side it had appeared on . However, there was a twist – unbeknownst to the mice, the game was “rigged” in their favor. In roughly 80% of the trials, the correct answer was, say, the left side. The mice learned this bias over time. So when the image was almost invisible, they would rely on that prior knowledge (i.e. “it’s probably on the left”) to make their choice . This ingenious design let scientists examine how the brain uses prior expectations to guide decisions when sensory evidence is uncertain.
Meanwhile, as the mice were making these left-or-right choices, the researchers were eavesdropping on their brains. Using nearly 700 ultra-thin electrode implants (called Neuropixels probes) inserted into the brain, the team recorded the electrical impulses of neurons in up to 279 different brain areas at once . All 12 labs adhered to the same procedures, so in total they recorded neural activity from 139 mice performing the identical task . The scale of data was staggering: about 620,000 neurons were monitored, covering approximately 95% of the mouse’s entire brain volume . In effect, the IBL scientists had wired up a mouse brain like a vast switchboard, capturing signals from nearly every region as the animal saw the cue, turned the wheel, and received feedback. Such a comprehensive recording had never been done in a mammalian brain before – until now, only much simpler creatures like fruit flies or tiny fish larvae had their whole brains monitored at this resolution . Little wonder that one neuroscientist called this achievement a “milestone… in terms of the type of specimen observed and the extent of the brain area covered.”
A Whole-Brain Decision Map
After crunching this mountain of data, the IBL team stitched together a dynamic portrait of brain activity unfolding during the decision process. The results were eye-opening. Instead of a single “decision center” lighting up, they saw activity rippling across the entire brain.
A brain-wide activity map of a mouse’s decision in action. Each colored dot represents a neuron firing as the mouse processes the visual cue, makes its choice, or receives a reward . The International Brain Laboratory combined recordings from tens of thousands of neurons to visualize how a decision engages circuits across the whole brain. Rather than one localized spot, multiple regions—from sensory areas at the back of the brain to frontal decision-making areas—light up together in a coordinated way. Researchers noted that the mouse brain’s activity during the choice (especially when the reward was delivered) lit up “like a Christmas tree,” reflecting how widespread the neural fireworks were .
This comprehensive map revealed that decision-making is a whole-brain affair, not confined to a few isolated areas. In classic neuroscience models, information was thought to flow in a hierarchy: for example, visual regions process an image, then higher-level cortical areas weigh the options, and finally a motor area triggers an action. But the new evidence challenges that traditional view . When the mouse was deciding, signals appeared in parallel across many regions – there was constant back-and-forth communication rather than a simple one-way chain . “Decision-making signals are surprisingly distributed across the brain, not localized to specific regions,” the researchers reported, emphasizing that even as the mouse initiated movement and received its reward, numerous areas were jointly active . In other words, making a decision looked less like a commander in one brain region issuing orders, and more like a symphony of brain areas performing together.
For example, when a faint image flashed on the screen, the mouse’s visual cortex (which processes sight) sparked with activity – but almost immediately, other areas jumped in too. That visual information spread in a wave-like pattern toward regions involved in evaluation and action . As the mouse committed to a choice and turned the wheel, neurons in motor regions, sensory regions, and reward centers were all buzzing in concert. One scientist described how the decision and reward events lit up the brain “like a Christmas tree,” with many bits of the brain glowing together . The take-home message is that even a simple decision engages a vast network: the brain works as an integrated whole, with far-flung regions acting in unison to drive the animal’s behavior .
The Brain’s Predictions Are Everywhere
Among the most intriguing insights from this study was how the brain handles prior expectations – the internal biases or “educated guesses” based on past experience. In the experiment’s hardest trials, the mice had to rely on memory of which side had been more frequent (since the cue was almost invisible). The neural data showed that the mice’s brains didn’t wait until the end to use that information – they wove it in from the very start. Expectation signals were found throughout the brain, even in areas that deal with early vision and basic sensory input .
In fact, the researchers were surprised to find signs of the mice’s expectations in the thalamus, a region that is essentially the first relay for visual information from the eyes . This means that the moment a mouse began a trial, its low-level sensory circuits were already “primed” with a guess about what it was likely to see. Higher-level decision-making regions (like parts of cortex) also carried these prior expectations, but the key revelation was that even the sensory-processing hubs carried a memory of the past . In the words of the scientists, the brain appears to act as a prediction machine, infusing expectations across multiple structures to help guide behavior . Rather than treating each new observation in isolation, the brain everywhere is subtly biased by what it expects will happen, based on recent experience.
This finding helps resolve a debate in neuroscience about how and where the brain integrates prior knowledge. Some theories held that early sensory areas simply feed raw data forward, and only higher cognitive areas mix in memories or context at a late stage. The IBL results point to a more pervasive mixing: even “early” regions are incorporating prior information from the get-go . For the mouse, knowing that the left side was likely proved useful, and its entire brain—from the visual thalamus up to decision-related cortex—reflected that bias as it decided . It’s as if the brain isn’t just a passive receiver of sensory input, but an active forecasting system that continuously predicts and updates what will happen next.
Understanding this brain-wide use of expectations isn’t only a win for basic science; it could also shed light on human health. Many brain and psychiatric disorders involve atypical decision-making or sensory processing, and researchers have speculated that problems with how the brain handles expectations (and surprise) might be a root cause. The new study’s authors noted that conditions like schizophrenia and autism might stem from differences in how expectations are formed or updated in the brain . Likewise, other experts have pointed out that disorders such as obsessive-compulsive disorder, Parkinson’s disease, and addiction – all of which can involve impaired decision-making – could be better understood by examining these widespread neural expectation signals . By providing a detailed map of how a healthy brain integrates prior knowledge to make a choice, the mouse brain study gives scientists a reference point for exploring what might go awry in these disorders.
Team Science and the Road Ahead
It took a globe-spanning effort to achieve this whole-brain window into decision-making, and the project may herald a new era of collaborative neuroscience. The International Brain Laboratory’s approach – multiple labs pooling expertise, standardizing their methods, and sharing data freely – was a deliberate departure from “business as usual” in brain research . “Traditionally, neuroscience has looked at brain regions in isolation. Recording the whole brain means we now have an opportunity to understand how all the pieces fit together,” noted Dr. Anne Churchland, a UCLA neurobiologist and core member of IBL . She called the outcome “a major step forward relative to the ‘piecemeal’ approach (1–2 brain areas at a time) that was previously the accepted method in the field.” To make sure their massive experiment was rock-solid, the team even formed a task force to enforce rigorous reproducibility across all participating labs – an important factor given the challenges of combining data from over a hundred mice and 12 research groups. The success of the IBL suggests that big, “team science” collaborations can answer brain questions that used to be out of reach for any single lab.
Crucially, all the data and tools from this project have been made openly available to scientists everywhere. The entire brain-wide dataset – comprising the recorded neural activity of those 620,000 neurons and the standardized experimental protocols – is publicly accessible on the IBL website for anyone in the neuroscience community to analyze . This means other researchers can dive into the data to ask new questions (for example, exploring one specific brain region’s role in the task, or applying AI algorithms to find hidden patterns). Already, the resource is being tapped: “The map is a fantastic resource that is already being mined by myriad scientists, and yielding unexpected discoveries. It’s a great success for team science and open science,” said Dr. Matteo Carandini, an IBL member at University College London . In short, the impact of this work goes beyond the initial Nature papers – it’s providing a treasure trove for the field, accelerating the pace at which new insights can be made.
And the work is far from over. Having demonstrated the power of a brain-wide approach to one type of behavior, the IBL plans to expand its scope beyond decision-making . Armed with new funding, the consortium aims to tackle other complex behaviors in the same collaborative, open-science spirit . The brain-wide map of decision activity, impressive as it is, is “a beginning, not the grand finale,” as one IBL scientist put it . The project has proven that herding dozens of neuroscientists to work together on a common goal can pay off in a big way. Now, the team envisions inviting more groups to join the effort, applying the IBL model to understand other mysteries of the mind .
By recording an entire mouse brain in the act of deciding, the IBL has delivered a striking confirmation that making a decision is never a local event – it’s a brain-wide symphony. The little mouse spinning its wheel in a lab “arcade” may seem worlds apart from a person choosing a latte or navigating a foggy road, but the fundamental lesson is likely the same. Our brains make decisions by engaging vast networks that integrate what we sense and what we expect. Thanks to this pioneering experiment, scientists have a richer map of that process than ever before . It’s a milestone that not only advances our understanding of the brain’s inner workings, but also highlights a path forward – one where neuroscientists join forces, pooling data and expertise, to chart the full complexity of the brain in action. In the grand quest to understand decision-making, the brain-wide map in mice is a giant leap, and it brings us a step closer to unraveling how our own minds weigh options and arrive at choices, one neuron (or half a million) at a time.
Sources: The International Brain Laboratory – UCLA Health News Release ; SingularityHub (Shelly Fan) ; Nature (Steinmetz et al., 2025) ; WIRED .