Brand New Study Detects Suicidal Thoughts with 90% Accuracy

At the present time, there is really no reliable way to realize whether a person is thinking about suicide. Even for very skilled professionals, it all comes down to guesswork and hunches. A therapist, if she suspects, could just ask her patient whether they are considering suicide.

The big problem is that most people are good at hiding it. Among the poor people who have actually committed suicide, 80% of them vehemently denied having any suicidal thoughts during their final visit with their mental health professionals. Because of these facts, scientists at Carnegie Mellon University began wondering whether they could create a way to detect these kinds of thoughts. They wondered if they could detect the thoughts through certain brain activity patterns.

Too Many Suicides among Young Adults

There is certainly a great need for this because about 44,000 Americans are committing suicide every year, and it is now the second highest cause of death for young adults. So if we had an objective way to detect these kinds of such thoughts, then perhaps we could create intervention methods that are more effective.

By using an MRI scanner and the proper machine learning algorithm, Carnegie Mellon University scientists think that they have isolated the brain signature that is associated with suicidal thoughts. The researchers discovered that they were able to predict the people who were having suicidal thoughts with 91% accuracy. They were also able to determine those that had previously attempted suicide from those who only thought about it. The study’s results were posted in the journal Nature Human Behavior.

Results Based on Small Sample Size

One thing that should be noted here is that these results came from a pretty small sample size. The study’s leaders feel that additional studies on a larger scale should be conducted to validate their findings. However, they are still compelling.

Researchers, who were led by Dr. Marcel Just, have gathered a significant number of neural signatures representing a variety of different emotions and thoughts. They are now able to determine what someone is feeling or even the type of social interaction they are currently thinking about, since each of these has its own unique thought pattern.

In this recent study, Just and his associates wanted to found out if certain thoughts were altered when a person contemplated suicide. They enlisted 34 volunteers and stuck them inside an MRI machine. Seventeen of these volunteers were picked because they had a history of having suicidal thoughts. And the remaining 17 acted as the control group.

The volunteers stayed in the machine for 30 minutes as life and death influenced words were displayed on a screen inside. The themes included words about good, death, praise, trouble, cruelty, and carefree. The negative verbiage was the focus as researchers believed these would elicit neural patterns that were associated with thoughts about suicide. Every word was displayed by itself for three seconds as the associated brain patterns were recorded by researchers.

Then all these MRI scans were submitted into the computer. A machine learning program evaluated the data and noticed differences between normal brain patterns and those that correlate with suicidal thoughts. After a bit of practice, this program became very skilled at distinguishing between the patterns. Those having suicidal thoughts registered different readings when the death-related words were displayed.