Predicting Individual Depression Symptoms from Acoustic Features During Speech

TL;DR

This paper explores using acoustic speech features with neural networks to predict individual depression symptoms, aiming to improve understanding and diagnosis of depression through detailed item-level analysis.

Contribution

It introduces a method to predict individual depression items from speech using CNN and LSTM models, incorporating temporal context and voting schemes for enhanced accuracy.

Findings

Neural networks can predict individual depression symptoms from speech.

Temporal context learning improves prediction accuracy.

Voting schemes influence the reliability of depression detection.

Abstract

Current automatic depression detection systems provide predictions directly without relying on the individual symptoms/items of depression as denoted in the clinical depression rating scales. In contrast, clinicians assess each item in the depression rating scale in a clinical setting, thus implicitly providing a more detailed rationale for a depression diagnosis. In this work, we make a first step towards using the acoustic features of speech to predict individual items of the depression rating scale before obtaining the final depression prediction. For this, we use convolutional (CNN) and recurrent (long short-term memory (LSTM)) neural networks. We consider different approaches to learning the temporal context of speech. Further, we analyze two variants of voting schemes for individual item prediction and depression detection. We also include an animated visualization that shows an example of item prediction over time as the speech progresses.