University Post
University of Copenhagen
Independent of management


AI can identify suicidal tendencies in young people

AI — Researchers can now spot behaviour showing who is at risk after building an algorithm and a huge data set. »We hope that our study can be used to predict and prevent a young person from ending up being in risk of suicide,« says associate professor of psychology.

How can a young person end up in a place where taking their own life feels like the only way out?

Are you having suicidal thoughts?

If you are in crisis or have thoughts of suicide, talk to someone. If you have acute thoughts of causing serious harm to yourself in Denmark, you should call 112.

Phone numbers for organizations that you can call (anonymously):

Startlinien – phone 35 36 26 00 – every day from 4 pm to 11 pm.

Sct. Nicolai Service – phone 70 120 110 – every day from 9 am to 3 am (except Sundays and holidays from 1 pm to 3 am.)

Sjælesorg – offers free online confidential chat on Mondays to Thursdays from 1 pm to 5 pm and from 7 pm to 10:30 pm.

Psychiatric Foundation – phone counseling 39 25 25 25. Monday to Friday from 11 am to 11 pm. Saturday, Sunday, and holidays from 11 am to 7 pm.

You can contact the phone counselling Livslinien on 70 201 201 every day of the year between 11 am and 5 am.

You can also write to the Livslinien email counselling, or contact Livslinien chat counselling.

A new collaboration between the University of Copenhagen and the University of Oslo will answer this question.

Using a purpose-built algorithm and a data set of 173,664 Norwegian young people aged 13-18, a team of researchers has mapped out the largest risk factors for suicide among young people.

»Previous studies have often focused on single factors, such as depression. But with the help of artificial intelligence, we have succeeded in finding out how different risk factors interact with each other,« says Milan Obaidi who is associate professor at the Department of Psychology, University of Copenhagen (UCPH) and is one of the project’s researchers.

In 2022, a survey from Aalborg University showed that up to every second young person is affected by unhappiness, according to the Danish news media DR. At the same time, the suicide rate in this age group is increasing at an alarming rate.

»We hope that our study can be used to predict and prevent a young person from ending up at risk of suicide,« says Milan Obaidi.

The biggest risk factor is recent self-harm

Milan Obaidi, associate professor at the Department of Psychology

Of the 173,664 young people taking part in the project Unveiling Adolescent Suicidality: Holistic Analysis of Protective and Risk Factors Using Multiple Machine Learning Algorithms, 4.65 percent have attempted suicide within the past 12 months. This corresponds to almost one in 20 young people.

Using the newly developed algorithm, the researchers have found a correlation among six risk factors, says Milan Obaidi.

»The biggest risk factor is recent self-harm. The other five risk factors are internalization problems, i.e. anxiety, depression, ADHD or similar diagnoses, sleep disturbances, eating disorders, hopelessness or lack of optimism about the future, and finally victimization, for example bullying,« he says and continues:

»The individual risk factors do not stand on their own. Of course you can eat poorly or have an actual eating disorder, without being suicidal. But according to the results of the project, the risk of suicide can arise when several of the factors are combined.«

The algorithm gets smarter

According to Milan Obaidi, artificial intelligence can help researchers spot psychological connections that conventional human analysis will not be able to see. In this specific experiment, the algorithm arrived at its results by comparing responses on state of mind and life circumstances among the many thousands of Norwegian young people.

»To my knowledge, no research has ever been done that combines the use of artificial intelligence with such large data sets,« he says.

We hope that our study can be used to predict and prevent a young person from ending up at risk of suicide,« says Milan Obaidi.

Milan Obaidi, associate professor at the Department of Psychology

»In psychology, you hardly ever work with such large amounts of data. Under normal circumstances, a study would be enormous if we had 500 participants. But with the help of artificial intelligence, it is possible to look at many variables at the same time.«

According to the researcher, the »beautiful« thing about the algorithm is that the more data it is fed with, the more accurate it will become over time.

»We don’t know exactly where the project will end, but the obvious thing to do would be to use the algorithm on a new corresponding data set next year, so we can monitor whether there are changes in the trends,« says Milan Obaidi.

Specifically, he hopes that municipalities or units that work with psychologically vulnerable and disadvantaged young people will be able to use the results to improve their preventive work with young people who are at risk of suicide.

»If you have a young person who is struggling, and you at the same time spot the six risk factors in the young person, we hope that you will be able to use that to take action as soon as possible,« says Obaidi.

In addition to himself, a number of researchers have been associated with the project – including computer scientists who have developed the algorithm.