After contributing work to a study on brain activity, Marcel Just described the effort to better understand human thought as “a psychologist’s dream.”

“Scientifically, it is fantastically interesting if you could assess the content of a thought,” Just, the director of the Center for Cognitive Brain Imaging at Carnegie Mellon, said.

With suicide being the second leading cause of death among college students in the United States, research by Drs. David Brent and Lisa Pan, professors in Pitt’s Department of Psychiatry, might be the start of better identifying who is at risk.

“It’s probably the most pressing clinical issue for a psychiatrist, is to determine somebody’s degree of risk,” Brent said.

The study — conducted by Brent, Pan and additional psychologists from Carnegie Mellon University, Florida International University, Columbia University and Harvard University — examined the association between brain activity in response to certain words and whether or not an individual has suicidal thoughts.

“You could argue that [suicide] is a biological experience and consequently, ‘not just in somebody’s head,’” Brent said.

Just, a collaborator in the study, said Brent came up to him after a talk he gave in September 2012 about brain activity and thought processing in individuals with autism.

“He said to me, ‘Do you think this would work on identifying suicidal thoughts?’ and I said, ‘Well we could find out with a research study,’” Just said. “And we were off to the races.”

The study consisted of 34 participants — 17 people who had suicidal thoughts and 17 who did not. Participants entered an fMRI, or functional magnetic resonance imaging, scanner where a screen within the scanner flashed a list of 30 words, one at a time for three seconds each. During this process, the fMRI detected which regions of the brain were activated.

To detect whether or not the person had suicidal thoughts, the researchers then processed the data they collected using a machine learning algorithm, which gives computers the ability to learn and process information without explicit programming.  

They fed the data on brain activity and suicidal thoughts into the machine learning algorithm for 33 of the 34 research participants. By looking at the final participant’s brain scan, the computer was instructed to follow the algorithm to determine whether or not that person has thought about suicide.

Researchers repeated the process so the computer had to guess whether or not each participant in the study had suicidal thoughts and it did so correctly 91 percent of the time.

“The most interesting thing, most important thing is possibly whether we can predict a suicide attempt,” Just said. “What would be extremely useful would be if we could predict if someone is likely to make a suicide attempt in the future. That could save lives.”

Six terms turned out to best distinguish those who had suicidal thoughts from those who did not — death, cruelty, trouble, carefree, good and praise. It was “maybe a little” shocking that “suicide” was not one of the six words from the original 30, Brent said, but the six words encompassed “words that were related” to suicide.

“We know that people have altered thinking about not just suicide but the words related to death and certainly words related to positive and negative emotions, they experience those words differently,” Brent said. “Afterward you can think about it, that a lot of people who are suicidal feel ashamed, they aren’t carefree.”

Madison Shaftic, a senior psychology major who struggles with two depressive disorders — persistent depression and major depressive disorder — as well as generalized anxiety, has struggled with suicidal thoughts in the past and discussed the stigma around the act.

“A lot of times I think there is a stigma with it in entertainment, specifically with TV and movies and all that, like they kinda romanticize it or they use it as a plot device and don’t actually treat it as a serious issue,” Shaftic said.

Brent said he didn’t believe the study would reduce stigma surrounding suicidal thoughts necessarily, but would open a dialogue with patients to discuss options for recovery.

“People make decisions to kill themselves on the basis of beliefs like ‘things aren’t gonna get better,’ ‘I’m a burden to people,’ and those things are not absolutely true,” Brent said. “Our job is to get people through a suicidal crisis so that they can realize that they could have a reasonable future.”