Scientists training computer programs based on brain scan to identify people with suicide

According to foreign media reports, scientists have trained computer programs based on brain scan to identify those who have suicide. The researchers said that this study is small, but this method may have been used for diagnosis of mental health. Nearly one million people around the world have died due to suicide, and they predict suicide is still difficult, especially because many people are discussing this problem is uncomfortable. In a study published in "Natural Communications" Monday, the researchers observed the brain activities of the two groups - a group of suicide people, a group of people without suicide ideas, they would think of "unfortunate" Or "praise" and other words. Researchers use these data to train one algorithm, the algorithm learns to predict the accuracy of 91%. The algorithm also predicts someone to try suicide with 94% accuracy. This algorithm is not perfect, medical test is different. It may not be widely used because the brain scan is expensive. However, Marchl Just, a psychologist at Carnegiemelon, believes: "This additional method is a good thing." 34 volunteers participated in this study: 17 suicide, 17 have not. Volunteers read 30 active ("Happiness", etc.) or negative ("cruel", etc.) or the word ("suicide") related to death ("suicide") and considering the meaning of brain scanning type known as FMRI. The researchers found that "death", "trouble", "carefree", "beautiful", "praise", "cruel" six words showed the biggest difference between the two groups of participants. They provide data to the algorithm 33 volunteers. For any given words, they tell which neural activation patterns come from which groups come from. They then provided the algorithm to the results of the last volunteer, and asked the algorithm to predict which group belonging to this person. The accuracy of the machine gives the answer reaches 91%. In the second experiment, scientists use the same method to teach an algorithm to distinguish between those who have not tried suicide, and this time the accuracy is 94%. The neuroscientist Blake Richards, the University of Toronto, said that the result is very interesting, but it may not be strong enough to make the test to diagnose. The activity mode is still correlation, not causality. He said: "This is a biological basis for someone who has to commit suicide. Every aspect of our spiritual life has a biological foundation, but the problem is whether the biological basis of these things can be fully utilized by FMRI to truly develop it. Reliable test of the clinical environment. " Original link: https://www.eeboard.com/news/intel-10nm-6/ Search for the panel network, pay attention, daily update development board, intelligent hardware, open source hardware, activity and other information can make you master. Recommended attention! [WeChat scanning picture can be directly paid] Related Reading: Guest officer, these 15 board cards are given to you, can you?