PsyDraw: A Multi-Agent Multimodal System for Mental Health Screening in Left-Behind Children

⚠️ Important Ethical Notice and Professional Use Guidelines

CRITICAL: This system is strictly designed as a professional screening aid tool.

This repository contains only the code structure of our project due to ethical considerations in mental health assessment. The system's prompts and analytical components are not open-sourced to prevent misuse and ensure proper application.

Professional Use Only

• This tool is exclusively designed to assist qualified mental health professionals (psychiatrists, psychological counselors, and school counselors) in preliminary screening.
• It MUST NOT be used for self-assessment or peer assessment by individuals without professional qualifications.
• All results require interpretation and validation by qualified mental health professionals.

Access to Full System

For research or clinical purposes requiring access to the complete system (including prompts), please contact us at: [[email protected]]. Access will only be granted after:

1. Verification of professional credentials
2. Review of intended use case
3. Agreement to ethical guidelines and usage terms

Risk Prevention

• Misuse of psychological assessment tools can lead to incorrect interpretations and potentially harmful outcomes
• The system should only be deployed in professional settings under qualified supervision
• All implementations must comply with relevant ethical guidelines and regulations in mental health assessment

Project Overview

Left-behind children (LBCs), numbering over 66 million in China, face severe mental health challenges due to parental migration for work. Early screening and identification of at-risk LBCs is crucial, yet challenging due to the severe shortage of mental health professionals, especially in rural areas. While the House-Tree-Person (HTP) test shows higher child participation rates, its requirement for expert interpretation limits its application in resource-scarce regions. To address this challenge, we propose PsyDraw, a multi-agent system based on Multimodal Large Language Models that assists mental health professionals in analyzing HTP drawings. The system employs specialized agents for feature extraction and psychological interpretation, operating in two stages: comprehensive feature analysis and professional report generation. Evaluation of HTP drawings from 290 primary school students reveals that 71.03% of the analyzes achieved High Consistency with professional evaluations, 26.21% Moderate Consistency and only 2.41% Low Consistency. The system identified 31.03% of cases requiring professional attention, demonstrating its effectiveness as a preliminary screening tool. Currently deployed in pilot schools, PsyDraw shows promise in supporting mental health professionals, particularly in resource-limited areas, while maintaining high professional standards in psychological assessment.

Figure 1: The workflow of PsyDraw

✨ Key Features

🚀 Quick Start

Installation

1. Clone the repository:

git clone https://github.com/LYiHub/psydraw.git
cd PsyDraw

2. Install dependencies:

pip install -r requirements.txt

3. Set up environment variables:

• Copy .env_example file and rename to .env
• Fill in your API key and base URL

Usage Methods

1. Direct Invocation

bash run.sh
# or
python run.py --image_file example/example1.png --save_path example/example1_result.json --language en

2. API Integration

python deploy.py --port 9557

Service runs on http://127.0.0.1:9557

3. Web Demo

bash web_demo.sh
# or
streamlit run src/main.py

4. Package Application

pyinstaller htp_analyzer.spec

📊 Case Studies

⚖️ License

This project is licensed under the GPL-3.0 License. See the LICENSE file for details.

⚠️ Disclaimer

PsyDraw is strictly a professional screening aid tool. It must not be used as a standalone diagnostic tool or a substitute for professional medical evaluation. The system is designed to support, not replace, the expertise of qualified mental health professionals. Any implementation or use of this system must be under professional supervision.

📚 Citation

If you find this work helpful, please cite our paper:

@misc{zhang2024psydrawmultiagentmultimodalmental,
      title={PsyDraw: A Multi-Agent Multimodal System for Mental Health Screening in Left-Behind Children}, 
      author={Yiqun Zhang and Xiaocui Yang and Xiaobai Li and Siyuan Yu and Yi Luan and Shi Feng and Daling Wang and Yifei Zhang},
      year={2024},
      eprint={2412.14769},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2412.14769}, 
}