TalkDep: Clinically Grounded LLM Personas for Conversation-Centric Depression Screening

TL;DR

This paper introduces TalkDep, a clinician-in-the-loop language model pipeline that generates diverse, clinically valid simulated patients for depression screening, aiming to enhance training and evaluation of diagnostic systems.

Contribution

We develop a novel simulation pipeline using advanced language models conditioned on clinical criteria, improving the realism and diversity of virtual depression patients.

Findings

Simulated patients are validated by clinical professionals for accuracy.

The approach enhances robustness of depression diagnosis models.

Provides scalable resource for training and evaluation.

Abstract

The increasing demand for mental health services has outpaced the availability of real training data to develop clinical professionals, leading to limited support for the diagnosis of depression. This shortage has motivated the development of simulated or virtual patients to assist in training and evaluation, but existing approaches often fail to generate clinically valid, natural, and diverse symptom presentations. In this work, we embrace the recent advanced language models as the backbone and propose a novel clinician-in-the-loop patient simulation pipeline, TalkDep, with access to diversified patient profiles to develop simulated patients. By conditioning the model on psychiatric diagnostic criteria, symptom severity scales, and contextual factors, our goal is to create authentic patient responses that can better support diagnostic model training and evaluation. We verify the reliability of these simulated patients with thorough assessments conducted by clinical professionals. The availability of validated simulated patients offers a scalable and adaptable resource for improving the robustness and generalisability of automatic depression diagnosis systems.