A study to determine the risk of suicide by combining brain scanning image and AI is in progress

What people think in the mind can not be seen from the outside, and even if the other person thought about "suicide", there are few clues to know about it . In order to solve such a problem, research is being carried out to combine fMRI , which visualizes the work of the brain, and image recognition technology by AI, to help identify the subject's "suicide risk".

A Brain Scan Could Help Reveal if a Person is a Suicide Risk | Digital Trends
https://www.digitaltrends.com/cool-tech/brain-scan-suicide-risk/

This research is being advanced by research teams at Carnegie Mellon University and the University of Pittsburgh. The research team has been awarded a grant of $ 3.8 million (about 420 million yen) from the National Institute of Mental Health (NIMH) to use state-of-the-art technology, and clinicians have signs of suicide We are engaged in research on technology to discover the subject and enable treatment.

What the research team focused on is the point of "the difference in brain activation pattern that occurs when people think about suicide and emotional concepts." People who tend to commit suicide and those who tend to suicide are death, cruelty, trouble, carefree, good, praise, "There are differences in the pattern of the parts activated in the head when you say the words, etc, and you measure the appearance with fMRI and use machine learning to judge suicide risk.

Although this research is still in its early stages, it is said that good results have been shown, such as succeeded in judging a person with suicidal risk with an accuracy of 91% using a machine learning algorithm (Gaussian Naive Bayes) .

Machine learning of neural representations of suicide and emotion concepts identifies suicidal youth | Nature Human Behavior
https://www.nature.com/articles/s41562-017-0234-y

Professor David Brent of the University of Pittsburgh at the University of Pittsburgh said, "The first study was relatively small, there were no subjects with no tendency to suicide, and it was cross-cutting to investigate the situation at one point in time." However, as for future prospects, "This research predicts future suicide risk by repeating the procedure to confirm whether fluctuation of neurosignal is related to fluctuation of suicidal tendency by further expanding the scale In addition, we aim to develop a method that can measure without using fMRI by developing and measuring a method that performs peripheral measurement. "

Once established, this method is expected to make it possible to judge the risk of suicide relied on subjects' self declaration in a more objective way. At the same time, by clarifying how people with suicidal risk are thinking about "death", it is expected to be utilized to lead psychotherapy more accurately.