Sign In

Communications of the ACM

ACM TechNews

Scientists Using Fitness Trackers, AI to Detect Depression with '80% Accuracy'

View as: Print Mobile App Share:

Computer scientist Iva Bojic said mental health disorders "can be visible in very physical ways," and wearable devices are helping them pick up those signals.

Credit: Matthew Ashe/Euronews

Scientists at Nanyang Technological University in Singapore (NTU Singapore) are tapping artificial intelligence (AI) to detect depression, schizophrenia, and other mental disorders earlier.

The researchers are developing a diagnostic toolkit to identify symptoms of such conditions, and to anticipate their worsening severity.

NTU Singapore's Iva Bojic said the team has collected "digital biomarkers," including heart rate and sleep pattern data, and correlated them with observed symptoms; this enables the AI model to learn.

Complementary questionnaires aid in the final assessment, with users in the sample classified as depressed or not.

The model can predict depression for new users based on their biomarkers, and Bojic estimated its accuracy at about 80%.

From EuroNews
View Full Article


Abstracts Copyright © 2022 SmithBucklin, Washington, DC, USA


No entries found