Surprisingly widespread brain activity supports economic decision-making, new study finds

A new study published in the Journal of Neuroscience sheds light on how the human brain makes decisions involving risk and reward. Researchers recorded direct electrical activity from multiple brain regions in neurosurgical patients and found that decision-making under uncertainty involves a widely distributed network of brain areas. While different regions contribute unique pieces of information—such as risk or win probability—the final choice emerges from the coordinated activity of many parts of the brain, especially through high-frequency signals that reflect active processing.

Previous research using brain imaging have often pointed to specific areas like the orbitofrontal cortex or striatum as central to evaluating rewards, findings from animal research have increasingly suggested a more distributed pattern. The current study took advantage of a rare opportunity to directly observe human brain activity with high spatial and temporal precision, using data collected from patients undergoing neurosurgery for epilepsy.

“In our daily life, we constantly make decisions that range from the trivial to the complex. However, even the easier decisions, such as what to have for lunch, require the coordinated activity of large areas of your brain,” said study author Ignacio Saez, director of the Laboratory for Human Neurophysiology and an associate professor at the Icahn School of Medicine at Mount Sinai.

“However, how these different areas act in a coordinated fashion to guide decisions has been difficult to pinpoint given how difficult it is to study the underlying biological activity, reflected by electrical activation, in the human brain. Here, we sought to better understand the neurobiological basis of economic decisions by carrying out electrophysiological recordings in multiple brain areas of patients undergoing brain surgery.”

The study involved 34 patients with drug-resistant epilepsy who had been implanted with electrodes in various parts of their brains to help locate seizure origins. While being monitored post-surgery, participants completed a simple gambling task. On each trial, they chose between a guaranteed monetary reward (a “safe bet”) or a riskier gamble offering a higher potential reward but with variable probability. The probability of winning was visually cued, and outcomes were revealed shortly after each choice. This task allowed researchers to isolate decision-making processes without requiring learning or memory.

From the initial group, 20 patients had behavioral and neural data of sufficient quality for analysis. Using intracranial electroencephalography (iEEG), the researchers recorded electrical signals from more than 1,000 electrodes placed in brain regions associated with decision-making, including the prefrontal cortex, motor and parietal areas, and deeper limbic structures such as the amygdala and hippocampus. They examined changes in brain activity across a range of frequency bands, from slow oscillations (like delta and theta) to high-frequency activity (gamma and high-frequency broadband activity, or HFA), which is thought to closely reflect local neuron firing.

The researchers found that during the moment of deliberation—just before a person made their choice—neural activity was modulated across a broad range of frequencies in many regions. However, the most consistent signals related to actual decision-making came from high-frequency activity. In contrast, slower brain waves seemed to reflect other processes, such as attention, goal orientation, or preparing for movement, depending on their location. For instance, increased theta and delta power were common in prefrontal and limbic areas, potentially signaling internal deliberation or memory retrieval, while beta decreases in motor areas may have reflected movement planning.

“In our study, we studied how players made choices between a safe bet and a risky gamble,” Saez told PsyPost. “We study multiple aspects of these decisions, including which types of information need to be considered (e.g. how likely uncertain choices are to result in a rewarding outcome and how risky they were), as well as the resulting final choice (safe bet or gamble). In this study, we surprisingly found that these types of information are much more widespread than previously thought, with many brain regions reflecting each of these different types of information, while maintaining some regional specificity (i.e. not all regions reflect all types of information equally strongly). This balance between regional specificity and global processing had not been previously shown in the human brain during decision-making.”

To better understand how brain signals were organized, the team grouped the observed regions into three broad functional circuits: prefrontal, frontoparietal, and limbic. Each circuit showed distinctive patterns of activity in both frequency and direction (increases or decreases in power), suggesting that different parts of the brain may support different cognitive components of the decision-making process.

Critically, the researchers found that high-frequency activity not only tracked the decision outcome (safe versus risky choice) but also encoded specific variables that inform the decision—such as the probability of winning and the amount of risk. These “choice-related computations” were distributed across many brain areas, but some regions were more involved in particular aspects. For example, the orbitofrontal cortex was especially sensitive to risk, while the postcentral gyrus, a region involved in motor control, represented which side of the screen the participant chose.

Interestingly, signals related to abstract variables like win probability and risk appeared earlier in time than the signal corresponding to the final decision itself. This timing supports the idea that the brain evaluates various components of a decision before settling on a choice. While these early computations were relatively localized, the final decision-related activity was more widespread, showing up across many regions in the form of increased high-frequency activity, especially when participants opted for the safer bet.

The study’s findings align with a growing view that the brain supports decision-making not through isolated modules, but through dynamic, overlapping systems. Unlike traditional brain imaging techniques that often highlight a few “hot spots,” this work demonstrates the value of direct recordings in uncovering more nuanced patterns of neural activity. The results suggest that while some regions may specialize in processing particular aspects of a decision, the final choice arises from a convergence of activity across the brain.

“The main takeaway is that your whole brain becomes active when you make an economic decision, from deep, evolutionarily old brain areas such as the amygdala to more recently evolved areas like the prefrontal cortex,” Saez explained. “Even though different brain areas are primarily implicated in different aspects of the decision (e.g. the orbitofrontal cortex, which has been long known to be implicated in decision-making, is the main area that represents the risk of the upcoming decision), electrical activity in all areas we recorded similarly reflected the nature of the choice (i.e. the participants’ choice). Therefore, we demonstrated that economic decision-making is a highly distributed process that does not uniquely depend on one or a few brain areas.”

As with all research, there are limitations to consider. The participants were epilepsy patients, and while their seizure origins were unrelated to the task at hand, their brain activity may differ in some respects from the general population. Electrode placement was dictated by clinical needs, so some brain regions were sampled more densely than others. Additionally, while the gambling task was useful for isolating key variables, it does not capture the full complexity of real-world decision-making.

“We carried out this research leveraging surgical interventions in epilepsy patients, which provide a unique opportunity to record electrophysiological activity from many brain areas with a high temporal resolution and anatomical precision that is otherwise not possible,” Saez noted. “This results in several caveats, the most important being that we did not have access to all brain areas, since the electrode location was determined on clinical, not research, grounds, and therefore we could not examine the contribution of some brain areas to decision-making. In addition, even though we found no behavioral differences with a healthy control population, all participants in this study were epilepsy patients.”

Despite these caveat, the study provides evidence that economic decision-making under uncertainty relies on a distributed network of brain regions. High-frequency neural activity appears to play a central role in encoding both the computations that guide choices and the decisions themselves. These findings bring new clarity to the neurophysiological basis of human choice and highlight the importance of studying brain function at fine temporal and spatial scales.

“Major psychiatric conditions such as depression or bipolar disorder are characterized by deficits in risky decision-making,” Saez explained. “In addition, new brain stimulation approaches are showing promise for developing neuromodulatory treatment strategies. Better understanding the neural basis of decision-making will allow us to further determine how this behavior is supported by brain activity, how this is affected in psychiatric disorders, and how to devise new neuromodulatory strategies for treating them.”

The study, “Distributed Intracranial Activity Underlying Human Decision-making Behavior,” was authored by Jacqueline A. Overton, Karen A. Moxon, Matthew P. Stickle, Logan M. Peters, Jack J. Lin, Edward F. Chang, Robert T. Knight, Ming Hsu, and Ignacio Saez.