A woman in glasses sitting in front of a bookcase.

U of A researchers detecting depression through voice

Health and Life Sciences
Published On
July 11, 2019
“I’m interested in applying and developing technology in the real world,” said Eleni Stroulia, professor of computing science at the University of Alberta. JOHN ULAN / SUPPLIED

Improved technology can now more accurately detect a depressed mood using just the sound of your voice, according to research by the University of Alberta.

“This is work that is important to us, that we are excited to do, that is rewarding and in the long term can be useful to humanity. But it’s too early, and we have to think about what’s the right way of making this happen,” said Eleni Stroulia, a professor in the department of computing science at the University of Alberta.

Doctoral student Mashrura Tasnim and Stroulia combined several algorithms, which are like recipes or “simple instructions to tell the computer what to do,” to recognize depression more accurately through voice cues, including volume, frequency and pitch, said Stroulia.

The goal is to develop a meaningful application from this technology that could help people understand depression. Approximately 11 per cent of Canadian men and 16 per cent of Canadian women will experience major depression in the course of their lives, according to the Government of Canada.

That could lead to the development of a mobile app, running on a user’s phone, that could collect voice samples, recognizing and tracking indicators of mood, much like step-counters that track physical movement, said Stroulia.

An app could be used by individuals tracking changes in their own moods over time, as a record for talking to a psychiatrist, or by caregivers. Data from voice cues could add more information to mental health assessments and treatment without relying on the patient to assess and record their emotional state.

“Yet another piece of information that hopefully helps diagnose things better,” said Stroulia.

But, there are legal, ethical and practical issues that would need to be addressed before an app helping to diagnose mental health issues would be useful. It could raise privacy concerns, so society has to think about how we deploy and manage these things, she said.

“This data cannot leak, but somebody has to access the data and make use of it. Absolute security guarantees uselessness,” said Stroulia. We need a legal and ethical framework to address data privacy issues, she added, especially since data is now part of almost all of our activities.

“Most disciplines collect data. We’ve had a technological advance that completely breaks open the types of data we can collect and the problems we could be addressing,” said Stroulia.