Speech-based Clinical Depression Screening: An Empirical Study

TL;DR

This study demonstrates that speech signals, especially deep speech features, are effective markers for AI-based depression screening across various interaction scenarios, with human-computer interactions matching clinical interviews in accuracy.

Contribution

It provides an empirical evaluation of speech-based depression screening, highlighting the effectiveness of deep speech features and the impact of interaction scenarios on model performance.

Findings

Deep speech features outperform traditional acoustic features.

Human-computer interaction matches clinical interview accuracy.

Segment duration and quantity influence model performance.

Abstract

This study investigates the utility of speech signals for AI-based depression screening across varied interaction scenarios, including psychiatric interviews, chatbot conversations, and text readings. Participants include depressed patients recruited from the outpatient clinics of Peking University Sixth Hospital and control group members from the community, all diagnosed by psychiatrists following standardized diagnostic protocols. We extracted acoustic and deep speech features from each participant's segmented recordings. Classifications were made using neural networks or SVMs, with aggregated clip outcomes determining final assessments. Our analysis across interaction scenarios, speech processing techniques, and feature types confirms speech as a crucial marker for depression screening. Specifically, human-computer interaction matches clinical interview efficacy, surpassing reading tasks. Segment duration and quantity significantly affect model performance, with deep speech features substantially outperforming traditional acoustic features.