Functional MRI Markers Measure Certainty — and Uncertainty — in the Brain of Normal Subjects

By Elizabeth Gardner

It makes sense that the level of uncertainty in a situation would change how a person's brain makes decisions about what to do next. Now researchers can actually see the differences, thanks to functional MRI (fMRI).

Farrar

On Monday, Danielle Farrar, MA, an MD/PhD candidate in the Boston University Department of Anatomy and Neurobiology, presented research showing that when neurologically normal subjects were certain in advance about the results of a decision, they used the areas of their brain most closely associated with rewards.

However, when they were uncertain, they used areas of their brain that are active during changing conditions, as well as areas associated with increased visual attention and motor control.

The goal of the study, which is part of Farrar's doctoral research, is to establish a baseline for how different decision-making tasks activate the brains of neurologically normal subjects who have not been diagnosed with any mental illness or neurological impairment.

With that baseline, Farrar eventually hopes to study impaired decision-making in such pathological states as schizophrenia, attention deficit/hyperactivity disorder, and obsessive-compulsive disorder. People with these conditions, as well as many other neuropsychological impairments, often have poor executive function, which includes not only decision-making, but the ability to plan, focus attention, remember instructions and follow them correctly, prioritize tasks, and control impulses.

"This work sheds light on the fact that executive function, thought to be a frontal lobe function, really taxes the whole brain during times of uncertainty," illustrating why those times can be so difficult for those with impaired executive function, Farrar said. Her work also seeks to establish how the various parts of the brain function in a network during the tasks studied, rather than simply studying the activity of individual areas.

Researchers examined 19 healthy subjects between ages 18 and 35 who each completed a decision-making card-matching task while undergoing an fMRI scan. The subjects were given one card, and asked to match it to one of five other cards. The card they were given matched each of the five cards in one attribute (such as color, number of shapes shown, type of shape shown, or type of border), but not in any of the others.

In the "certain" condition of the card-matching task, the subject chose a rule to apply for matching the first card, and that rule remained in effect for all of the "certain" trials.

In the "uncertain" condition, a program randomly chose the correct rule for each trial and the subject didn't know it in advance, so the likelihood of making an incorrect choice was high.

Each subject received a baseline structural scan and then completed the matching tasks six times, alternating with brief rests to return to baseline, during a 36-minute functional scan. The fMRI scan captured brain function every two seconds. Cards were presented for four seconds and subjects were instructed to respond within that time on a button box held in their right hand. Certain and uncertain conditions alternated. Half the subjects started with the certain condition and the other half started with the uncertain condition. The results showed different networks of functional brain regions for the two conditions.

The areas activated in the "certain" condition — the insula, parietal cortex, temporal cortex, ventromedial cortex, and orbitofrontal cortex — were those generally associated with reward. The "uncertain" condition activated areas usually associated with resolving uncertainty and updating rules (the prefrontal cortex, parietal cortex striatum, thalamus, amygdala and hippocampus), and also activated the occipital cortex and midbrain, areas associated with visual attention and motor control.