AI capable of accurately predicting suicide risks

Artificial Intelligence could hold the answer to predict suicide risks and prevent the tragedy. Learning machines were able to correctly assess risk situations and provide insight into how suicide attempt risk shifts over time.

Close to 800,000 people die due to suicide every year and many more attempt suicide. It is a global phenomenon that affects every region and it is the second leading cause of death among 15-29-year-olds globally.This makes correct predictions of suicide risks extremely important as a tool to prevent tragedy and psychologists are now looking at the recent technological advancements in order to see if machines could help drive down the numbers of people that die as a result of self-harm.

Researchers from the University of Florida led a groundbreaking project that suggests that machine learning could improve suicide risk prediction. By how much? According to their results, clinicians could have two years prior notice and an accuracy of 80 per cent.

FSU Psychology researcher Jessica Ribeiro’s paper, published in the Journal of Clinical Psychological Science is titled “Predicting Risk of Suicide Attempts over Time through Machine Learning“. In it, scientists detail their findings that show artificial intelligence could help predict, with 80 to 90 per cent accuracy, the risk of suicide.

What is fascinating about the results is not only their accuracy but also the fact that they are able to predict with two years in advance if someone will attempt suicide. And the algorithms become even more accurate as a person’s suicide attempt gets closer.

For example, the accuracy climbs to 92 per cent one week before a suicide attempt when artificial intelligence focuses on general hospital patients.

“This study provides evidence that we can predict suicide attempts accurately,” Ribeiro said in an article for the University. “We can predict them accurately over time, but we’re best at predicting them closer to the event.”

Risk factors play different roles and change over time

The study also offered insight into how different risk factors play a part in driving someone to attempt self-harm. This is especially important as looking at traditional factors like depression, stress or substance abuse did not provide a reliable measuring unit for predicting suicide.

“We also know, based on this study, that risk factors, how they work and how important they are, also change over time,” added Ribeiro.

Ribeiro previously worked on a study that concluded that 50 years of suicide prediction research had not produced any real progress in being able to predict who will try to kill themselves. But the access to the health data base it provided had proved crucial in teaching the machine the combination of risks that can trigger suicide attempts.

The team combed through the electronic health records, which were anonymous, and identified more than 3,200 people who had attempted suicide. The documents contained detailed medical histories of thousands of people leading up to their suicide attempts. This is how the machine learned what were the signs that lead to suicide attempts.

“The machine learns the optimal combination of risk factors,” Ribeiro said. “What really matters is how this algorithm and these variables interact with one another as a whole. This kind of work lets us apply algorithms that can consider hundreds of data points in someone’s medical record and potentially reduce them to clinically meaningful information.”

“Red light warning” to save lives

Using these algorithms, clinicians could receive a prior notice, helping them prevent the suicide. Scientists believe that their findings could lead to the creation of a system that will issue a “red light warning” or “risk score” for people coming into a hospital with acute symptoms. This way an emergency room doctor could see an elevated risk for suicide and get a psychiatrist to respond immediately.

“Just like you get a cardiovascular risk score, you would get a suicide risk score that is informative for clinicians and helps direct them on what steps to take next,” Ribeiro said.

Ribeiro was prompted in her study by the grim statistics that show that 120 Americans take their lives every day and according to health data, 60 to 90 per cent of them visited their medical provider within the past year and the clinician never saw it coming.

“This algorithm funnels our attention to the folks who are most likely to attempt suicide so our resources are better devoted to people we’re missing now,” Ribeiro said. “Right now, we’re missing a large proportion of people who are at risk that we never even think about.”

Ribeiro is currently working on a study with the Military Suicide Research Consortium based at Florida State that uses machine learning to identify people with an imminent risk of suicide.