Unveiling the Landscape of Clinical Depression Assessment: From Behavioral Signatures to Psychiatric Reasoning

TL;DR

This paper introduces C-MIND, a comprehensive clinical depression dataset, analyzes behavioral and multimodal data for diagnosis, and explores the capabilities and limitations of large language models in psychiatric reasoning.

Contribution

The paper presents C-MIND, a new multimodal dataset collected from real clinical settings, and evaluates classical models and LLMs for depression diagnosis, proposing methods to improve LLM reasoning with clinical guidance.

Findings

Classical models show varying effectiveness across tasks and modalities.

LLMs have limitations in psychiatric reasoning within clinical contexts.

Guided reasoning with clinical expertise improves LLM diagnostic accuracy by up to 10%.

Abstract

Depression is a widespread mental disorder that affects millions worldwide. While automated depression assessment shows promise, most studies rely on limited or non-clinically validated data, and often prioritize complex model design over real-world effectiveness. In this paper, we aim to unveil the landscape of clinical depression assessment. We introduce C-MIND, a clinical neuropsychiatric multimodal diagnosis dataset collected over two years from real hospital visits. Each participant completes three structured psychiatric tasks and receives a final diagnosis from expert clinicians, with informative audio, video, transcript, and functional near-infrared spectroscopy (fNIRS) signals recorded. Using C-MIND, we first analyze behavioral signatures relevant to diagnosis. We train a range of classical models to quantify how different tasks and modalities contribute to diagnostic performance, and dissect the effectiveness of their combinations. We then explore whether LLMs can perform psychiatric reasoning like clinicians and identify their clear limitations in realistic clinical settings. In response, we propose to guide the reasoning process with clinical expertise and consistently improves LLM diagnostic performance by up to 10% in Macro-F1 score. We aim to build an infrastructure for clinical depression assessment from both data and algorithmic perspectives, enabling C-MIND to facilitate grounded and reliable research for mental healthcare.