It is difficult to determine whether a person is experiencing suicidal tendencies simply by asking them about it. According to research, 80% of those experiencing this threshold deny the desire to commit suicide.
Scientists have created AI technology that can identify this fact, and, perhaps, prompt intervention from specialists. They developed a machine learning algorithm that determines suicidal tendencies based on the activity of certain areas of the brain. Previously, the researchers invited volunteers who had ever tried or planned to take their own lives. During the experiment, they thought about words associated with suicide, negativity, or those relating to the opposite positive emotions.
So, based on data from the fMRI algorithm, the team identified areas of the brain that were activated when words associated with suicide were mentioned. The system coped with the task by 91%, separating people from the risk group. Scientists warn that the conclusions drawn by the current method cannot absolutely guarantee the authenticity of the fact. The studies were conducted on volunteers who flawlessly followed all instructions; in real life, a person will want to deceive the system, focusing on other things. Plus, the conclusions were drawn from 30-minute sessions using giant MRI scanners, which are completely impractical for daily use. However, the likelihood that in the future AI will be able to identify such signals increases significantly.
GO TO FULL VERSION