Investigating the Physiology of Human Decision-Making

Lay Summary

Deciphering the physiology of the human decision-making network

Neurosurgical researchers will explore how a specific brain network functions as people make decisions.

We are constantly making decisions. Often we make them rapidly with nearly imperceptible effort. Other decisions require us to make a thorough but rapid examination of the circumstance, predict potential consequences and take action. Sometimes these actions are essential for survival and it is no surprise that the specialized networks for making decisions are located in the brain’s most evolutionarily advanced region, the pre-frontal cortex. Take, for example, deciding whether to stop at or drive through a yellow light. This situational context creates conflict between multiple simultaneous and mutually incompatible responses: depress the brake, step on the accelerator, or coast. The networks involved evaluate these options taking into account multiple factors, such as the car’s speed, distance to the intersection, and whether other cars (or police) are around.

Scientists have postulated that the decision process entails two functions. One function entails evaluating the situation by identifying the sources of conflict, while the other entails exerting control by initiating actions to resolve the conflict and execute the optimal response. Prior research suggests that an area in the middle of the pre-frontal cortex, called the dorsal anterior cingulate cortex (dACC), is primarily involved in monitoring conflicts while the dorsolateral pre-frontal cortex (DLPFC) exerts cognitive control to resolve conflict and optimize behavior.

Previous MRI studies suggest that the dACC monitors conflict and evaluates possible responses for their likely outcomes, and then sends signals to the DLPFC. When the demands are relatively stable, such as in deciding which suit to wear, this signal accelerates responses and promotes efficiency. In contrast, when demands are rapidly changing—as in the yellow light example—the signals retard responses and promote accuracy. In other words, in critical decision-making situations, accuracy trumps efficiency.

In this model, the dACC recruits the DLPFC to make adjustments in cognitive control. Investigation of the physiological process of cognitive control hinges on two behavioral observations. The first observation is that our reaction times are longer when there is incongruence (“cognitive interference”) between relevant and irrelevant stimuli, such as when the word “red” is written in green ink compared to when it is written in red ink. There is a tendency for the irrelevant green ink to impede our simultaneous processing of the relevant word “red”. The second observation is that we tend to adjust our current response based on recent past occurrences.

The Columbia University neurosurgical researchers are among the few groups in the country with the expertise and experience to study neural networks from the levels of individual cells and populations of cells. These investigators will extend prior research to answer key questions about the roles of dACC and DLPFC circuitry in cognitive control. They will study this decision-making network in participants as they make decisions on tasks that entail cognitive interference or recollection of recent prior experience. Investigators will compare the results from electrical recordings of single brain cells and of populations of cells to results obtained from fMRI imaging. The studies are designed to bridge the gap between findings from imaging and cellular recording and from human and non-human primates.

One study group will involve 30 patients with intractable epilepsy who will participate in the decision-making tasks prior to their surgical treatment. These patients ordinarily undergo electrode recordings prior to surgery so that the surgeons can identify brain areas that are critical for specific functions and need to be surgically sparred. As these patients participate in the two behavioral tasks, the surgeons will study the dACC to test three hypotheses: 1) The dACC, by weighing the most recent events most heavily, helps guides immediate decisions; but, by taking into account previous events, it remains sensitive to the historical context. 2) The dACC neurons encode possible responses that are activated by the amount of conflict. 3) The discrepancy in prior results from single cell responses and fMRI imaging will dissolve when the entire time-course of the trial is taken into account.

The other study group will consist of 30-45 patients with either intractable Parkinson’s disease or the related condition, “essential tremor.” The investigators will similarly undertake fMRI imaging and single and multiple cell electrode recordings in patients prior to undergoing deep brain stimulation (DBS) surgery. The neurosurgeons will first describe basic response properties of DLPFC neurons, and compare results to those previously obtained in non-human primates. Then they will test the hypothesis that the dACC signal recruits DLFPC control mechanisms to contend with conflict and optimize behavior.

Significance: This research is anticipated to provide a new understanding of the physiology of the network involved in normal decision-making. This understanding might then also shed light on disorders thought to arise from physiological malfunctioning of this network, such as obsessive-compulsive disorder, attention deficit hyperactivity disorder and schizophrenia