Voice tracking app could detect depression, scientists say

Scientists have revealed they're planning to create a smartphone app that detects if someone's depressed based on changes in their voice. Speech coordination changes when a person becomes depressed, according to the researchers, at the University of Maryland. Depressed people cannot think as fast, and their speaking rate is slowed with more and longer pauses than if they are not depressed, they say. Therefore, a voice detection app using deep learning – a type of machine learning based on artificial neural networks – could help detect such traits, which can often be subtle. The app could be recommended by mental health therapists to their patients, who would submit video and audio updates on their mood at home, which the technology would then assess.