It's difficult to identify whether or not someone is feeling suicidal by asking. According to studies, roughly 80 percent of those who go through with the act denied feeling suicidal the last time they spoke to a mental health care expert. However, AI-related technology -- which is also being applied to things like spotting cancerous cells -- might offer a way to identify these thoughts and prompt a timely intervention. Researchers have developed a machine learning algorithm that identifies suicidal tendencies based on activity in specific brain regions.
Based on fMRI scans, the team learned that certain keywords sparked activity in mostly different brain regions depending on whether or not someone was feeling suicidal. With that knowledge, the team produced an algorithm that can measure the overall brain activity and determine when activity in suicide-linked regions is particularly strong. The system is 91 percent accurate in tests and is as trustworthy in identifying unaffected people as those with suicidal thoughts.
Lead author Dr. Marcel Just was quick to warnThe Methods Man that this isn't a guaranteed method for revealing suicidal thoughts, at least not in its present form. The conclusions came from 30-minute sessions in giant MRI scanners, which are neither practical for therapists' offices nor particularly inviting. Also, everyone in the study was a volunteer. It remains to be seen whether or not a suicidal person who hasn't confessed their feelings will produce the same results. If they know doctors are looking for specific brain activity, they may suppress their reactions in a bid to avoid divulging their true thoughts. Even so, this raises the possibility that AI could reveal these cues in the future by using less conspicuous equipment and testing methods.