Suicide is the second leading cause of death among young people aged 15-34 in the United States, and clinicians can only use limited means to identify those who have suicidal tendencies. Today, published in the "natural human Behavior "in a paper that records a new machine learning technique that can help identify people who have suicidal thoughts.
The investigators conducted a survey of 34 young people with half of the suicidal participants and the experimental control group, each of which was subjected to functional magnetic resonance imaging (fMRI) and gave three samples containing 10 (Such as 'death', 'pain' or 'fatal'), positive influence ('carefree', 'good', 'naive') or Negative effects ('boring', 'evil', 'guilt'). The researchers also used the previously drawn neural signal maps that show the emotional patterns of the brain, such as 'shame' and 'anger'.
Five brain targets, together with six words, were found to be the best signs of distinguishing between suicide and control groups, and using these positions and vocabulary, the researchers trained a machine learning classifier that correctly identified 17 suicides Of the 15 suicidal participants, and 16 of the 16 control subjects without suicide tendencies.
Subsequently, the researchers divided the suicide patients into two groups, one had suicide experience (9), the other had no suicide experience (eight), and trained a new classifier, which was able to correctly identify 17 16 patients in the patient.
The results showed that participants in mental health and those who had suicidal thoughts responded differently to words, for example, when suicidal participants saw the word 'death', their 'shame' area of the brain was larger than that of the control group In the same way, the word 'trouble' also triggers more brain activity in 'sad' areas.
This is the latest attempt to bring artificial intelligence to psychiatry, and researchers are studying machine learning projects ranging from analyzing NMR spectra to predicting major depressive disorder to identifying PTSD from people's patterns of speech, .
Earlier this year, Wired magazine reported that some researchers had established a system that could record the risk of suicide by analyzing health records, with an accuracy rate of 80% to 90%. Facebook was in use Text mining techniques to identify users who face the risk of suicide or crippling, and then point them to mental health resources (see 'Facebook's suicide prevention tools' issues].
Artificial intelligence has set off waves in the medical field, and some algorithms are very good at detecting other problems in tumors and CT scans, and Jeffrey Newton tells the "New Yorker" that the radiologist will eventually be unemployed. One of the most important researchers in the field of study.In fact, he said, 'They should now stop developing radiologists.'
In this case, the study is more likely to stimulate new human-driven therapies rather than letting doctors in the whole field lose their jobs.This paper points out that identifying different patterns and regions can be developed for brain stimulation New areas. Specific emotional responses to suicide-related terms can also be used in psychotherapists to treat patients.