Innovative Framework for Early Estimation of Mental Disorder Scores to Enable Timely Interventions

TL;DR

This paper introduces a multimodal deep learning framework combining textual and audio data to accurately classify depression and PTSD, enabling earlier detection and intervention in mental health care.

Contribution

The study presents a novel multimodal deep learning approach that integrates LSTM and BiLSTM architectures for improved mental disorder classification from clinical interview data.

Findings

Achieved 92% accuracy for depression detection.

Achieved 93% accuracy for PTSD detection.

Outperformed unimodal approaches in classification accuracy.

Abstract

Individual's general well-being is greatly impacted by mental health conditions including depression and Post-Traumatic Stress Disorder (PTSD), underscoring the importance of early detection and precise diagnosis in order to facilitate prompt clinical intervention. An advanced multimodal deep learning system for the automated classification of PTSD and depression is presented in this paper. Utilizing textual and audio data from clinical interview datasets, the method combines features taken from both modalities by combining the architectures of LSTM (Long Short Term Memory) and BiLSTM (Bidirectional Long Short-Term Memory).Although text features focus on speech's semantic and grammatical components; audio features capture vocal traits including rhythm, tone, and pitch. This combination of modalities enhances the model's capacity to identify minute patterns connected to mental health conditions. Using test datasets, the proposed method achieves classification accuracies of 92% for depression and 93% for PTSD, outperforming traditional unimodal approaches and demonstrating its accuracy and robustness.