How are you feeling? The clues to your mental health might be hidden in your phone.
Ubiquitous in people’s lives, cellphones and what they do on them can be a window into who they are — tracing feelings, desires and insecurities. In an era where much of what people do leaves a digital footprint, UO psychology researcher Nick Allen is developing an app that will use this information to predict suicide risk and potentially save lives.
“What clinicians and therapists really need is some way of knowing not just who is likely to attempt suicide or have suicidal thoughts and behaviors, but when they might do it,” said Allen, a professor in the Department of Psychology.
Allen directs the Center for Digital Mental Health, a research institute that develops ways to use digital technology to improve mental health. Supported by a $2.9 million grant from the National Institutes of Health, this four-year, UO-led project will involve a partnership with mental health researchers, clinicians and computer scientists at Columbia University, University of Pittsburgh and Carnegie Mellon University.
In this latest research, the center will dive deep into the cellphone habits of a tech-saturated population: teenagers. The aim is to use this cellphone data to uncover patterns that can identify the warning signs for those at risk of suicide.
According to Allen, one of the greatest challenges in the area of suicide prevention is the absence of a short-term method of prediction.
“For most people that have suicidal intent, the actual period where they’re likely to act is only about 10 to 15 minutes,” Allen said. “If you can help get someone through that high-risk period, often they will reach out for help or they will do something different.”
Researchers will focus on 200 youths age 13 to 18 who have attempted suicide or had suicidal thoughts. Patients at the University of Pittsburgh and Columbia University clinics will have an intake assessment and then install the application on their phone, where it will track their activities for a period of six months.
Researchers will have an intimate window into the lives of the participants, following their day-to-day activities such as geographic movement, physical activity, sleep patterns, what kind of music they listen to, who their social networks are and the kinds of messages they’re sending on social networks.
“We can quantify a huge amount of day-to-day behavior from their natural phone usage
without us even having to ask them a single question,” Allen said.
Periodically, researchers will send push notifications with survey questions for participants that will inform the interpretation of the data. The app will even use new technology to track facial expressions and aspects of their voice and speech to help decode their emotional state.
“We can often say, Iin general, this person is at (greater) risk, than another person,’” Allen said. “But (without this technology) the time lag between identifying that risk and the potential suicidal event can be years.”
Once the data is collected, the UO will work with machine learning specialists at Carnegie Mellon University to develop algorithms that will identify indicators of suicidal risk. The project has the potential to revolutionize how mental research and interventions are conducted.
“Ultimately, we hope that (this) is going to allow us to offer interventions to people exactly when they need them,” Allen said. “And when they’re likely to be the most effective.”
—By Piper McDaniel, Office for Research and Innovation