As they watched 24 people play games of rock-paper-scissors, Yale researchers found that their minds were broadly focused on winning.
A research team, led by Yale psychology postdoctoral fellow Timothy Vickery and two other Yale professors, has discovered that more regions of the brain are involved in processing rewards than scientists previously thought. Their report, published in the Oct. 6 issue of the journal “Neuron,” suggests that when humans play games, wins and losses engage parts of the brain beyond reward centers in the basal ganglia.
In the study, human subjects were given financial incentives to win many rounds of simple games such as rock-paper-scissors and matching pennies, which give each player an equal chance of winning. $4 was awarded to subjects who won a 53-trial round of rock-paper scissors. While playing, the researchers scanned the participants’ brains using a common technique called functional magnetic resonance imaging, or fMRI.
“Traditional ways of analyzing brain scans could only detect a few areas involved in processing wins and losses,” Vickery said. “I wanted to see what other regions could be involved.”
While most studies on the reward system analyze blood oxygen levels to determine the amount of brain activity, Vickery said he examined patterns in blood flow. Traditional analytical techniques only detect differences in activity levels, he added.
“Other regions that are involved might be missed since the reward information might not be coded in activity level but rather in patterns,” he said.
By training a computer program to distinguish shifting brain activity patterns in the fMRI scan, Vickery found that 37 out of 43 regions of the brain showed patterns corresponding to wins and losses for players of matching pennies, compared to around eight regions found using traditional analytical techniques.
When he further divided the brain into over 270,000 areas, over 30 percent of areas showed a significant response, compared to only eight percent detected with other methods. Similar though less extreme discrepancies were found in the game of rock-paper-scissors.
“Just by examining the patterns in most areas of the brain, we can show when a win or loss has resulted,” Vickery said.
Michael Anderson, an assistant professor of cognitive science professor at Franklin and Marshall College, pointed out that the signal strengths were relatively weak, making it hard to identify with certainty which regions of the brain were involved. The pattern recognition algorithm was only able to classify individual regions as involved or not at a slightly better rate than random chance would have, Anderson said.
Although the results are still significant, he said researchers will need to address how these weak signals all play a role in the reward system.
Vickery said that the challenge for researchers going forward is to determine the underlying cause of the results and determine their role in reward processing.
“What’s being actually coded?” Vickery said. “And are the signals there for learning in those domains or do they have some other role?”
Roy Wise, a senior investigator at the National Institute on Drug Abuse, said the study’s results support existing theories associating reward processes that contribute to learning with many different parts of the brain. For example, Kent Berridge, professor of psychology and neuroscience at the University of Michigan said that some of the signals might come from thinking about the emotions of winning or losing or remembering past instances of those emotions.
Vickery said he thinks there’s a widening agreement among psychologists that reward signals are distributed throughout the brain, adding that he hopes further studies will determine the role of less central parts of the brain in the reward system.
Psychology professor Marvin Chun and associate professor of neurobiology and psychology Daeyeol Lee assisted in the study.