AgentMental: An Interactive Multi-Agent Framework for Explainable and Adaptive Mental Health Assessment

TL;DR

This paper introduces an interactive multi-agent framework for mental health assessment that simulates clinical dialogues, employs adaptive questioning, and uses dynamic memory to improve information extraction and diagnostic accuracy.

Contribution

It presents a novel multi-agent system with adaptive questioning and tree-structured memory for more effective and explainable mental health evaluation.

Findings

Outperforms existing methods on DAIC-WOZ dataset

Enhances information extraction through dynamic memory updates

Improves diagnostic accuracy with adaptive questioning

Abstract

Mental health assessment is crucial for early intervention and effective treatment, yet traditional clinician-based approaches are limited by the shortage of qualified professionals. Recent advances in artificial intelligence have sparked growing interest in automated psychological assessment, yet most existing approaches are constrained by their reliance on static text analysis, limiting their ability to capture deeper and more informative insights that emerge through dynamic interaction and iterative questioning. Therefore, in this paper, we propose a multi-agent framework for mental health evaluation that simulates clinical doctor-patient dialogues, with specialized agents assigned to questioning, adequacy evaluation, scoring, and updating. We introduce an adaptive questioning mechanism in which an evaluation agent assesses the adequacy of user responses to determine the necessity of generating targeted follow-up queries to address ambiguity and missing information. Additionally, we employ a tree-structured memory in which the root node encodes the user's basic information, while child nodes (e.g., topic and statement) organize key information according to distinct symptom categories and interaction turns. This memory is dynamically updated throughout the interaction to reduce redundant questioning and further enhance the information extraction and contextual tracking capabilities. Experimental results on the DAIC-WOZ dataset illustrate the effectiveness of our proposed method, which achieves better performance than existing approaches.