The Sound of Depression Captured By Mobile App


Thanks to a new research coming from the University of Alberta, depression can now be spotted by artificial intelligence that identifies the characteristic sounds of grief in a person’s voice accurately.

The research started from a past trial, which involved that the timbre of voice carries valuable information regarding the holder’s mental state. The software warns when the speech matches the voice patterns of depressed individuals.

The vocal trait involved is also a ‘paraverbal feature.’ Some of the qualities of depressed speeching patterns can be noticed only by a trained ear, while others are more subtle. Recently, with the help of technology, the changings in tone are picked up and quantified.

The Mobile App That Stands In for a Psychiatrist

The Statistics of the Federal Government show that the incidence of depression is also high in Canada – 16 percent of women and 11 percent of men will encounter a major episode of depression at least once in their life. Less than half of those suffering receive specialised treatment.

The teenagers with ages between 12 and 19 are also a vulnerable category to psychiatric disorders, says Canadian Mental Health Association.

In an attempt to fill that gap, the modern world is creating digital technology that flags depression.

It’s Not About What You Say, but How You Say It

The approach contains several machine-learning algorithms for a more reliable detection, based on acoustic cues. It was improved by Ph.D. student Mashrura Tasnim and professor Eleni Stroulia.

The technology was projected in the first place as a tool for everyone. It is perfect for care providers to alert about sudden changes in behavior.

Much like you have a step counter on your phone, you could have a depression indicator based on your voice as you use the phone – she said.

The app will record phone users while they speak naturally and draw up mood reports, such as depression levels. The voice sample and data are at a safe place, locked in an user account.


Related Posts