Crisis text line, suicide prevention through artificial intelligence

Publications Bank of innovations

Crisis text line, suicide prevention through artificial intelligence

Crisis Text Line

Suicide prevention hotline which, in addition to offering accompaniment, emotional support and home interventions to thousands of people going through difficult situations, has an algorithm that allows the service’s volunteers to identify which people are at high risk of suicide and therefore need more urgent intervention.

The service works in such a way that the person in crisis sends a message saying “HOME” to the number 741741, and a specialised volunteer responds quickly and in real time. The service goes out of its way to help people in times of distress or crisis by listening to them, reassuring them, and offering suggestions for overcoming their difficulties.

As a service that receives thousands of requests, deciding who to help first in times of floods of messages can be a life-or-death decision. That’s why Crisis Text Line uses an algorithm to identify people at high risk of suicide and to be able to intervene as quickly as possible in these cases.

Using machine learning, the artificial intelligence algorithm analyses the words and emojis that may indicate that a person is at high risk of suicidal ideation or self-harm. The system developed is based on the analysis of 30 million messages that the service has exchanged with users in recent years, and is thus able to position as an emergency those conversations with high-risk patterns that need to be attended to more urgently. As a result, 94% of people at serious and imminent risk of suicide receive an intervention in less than 5 minutes.

In addition, the data collected provides the service with very relevant information about the mental health of the clients. For example, it has concluded that Wednesday is the most anxiety provoking day of the week, and that self-harm crises occur mainly in the early hours of the morning.

Banc d’innovacions

Energy Control

Energy Control, a drug analysis service to reduce risks

Energy Control is a drug analysis program that promotes responsible consumption
BetiON

Predictive Telecare Service in Home Care

The Adinberri Foundation has launched a telecare service that enables anticipation and action through a personalized intervention plan based on the 4P model: prediction, prevention, personalization, and participation.
From Your Eyes

From your eyes, AI technological tools for visually impaired people

From Your Eyes promotes AI-based solutions to improve the daily lives of visually impaired people.
Reviu

Reviu, a platform that promotes transparency in housing access

Reviu is an online platform for sharing and searching for apartment reviews, with the primary goal of providing transparent and real information about the real estate market.
I·ROC

I·ROC, a tool that evaluates the recovery process of mental health service users

I·ROC is a tool that measures the recovery process of people using mental health services
SmartSocks

SmartSocks, socks that detect distress in people with dementia

SmartSocks that enable the detection of risk situations in people with dementia