Home // eTELEMED 2025, The Seventeenth International Conference on eHealth, Telemedicine, and Social Medicine // View article

AI-Assisted UML Modeling for Serious Mental Illness Crisis Management

Authors:
Kerry Gilder
John Chelsom

Keywords: artificial intelligence; UML diagrams; serious mental illness; PlantUML; AI-assisted modeling.

Abstract:
This study assesses the effectiveness of Artificial Intelligence (AI) generated Unified Modeling Language (UML) diagrams in illustrating the treatment pathways for Serious Mental Illness (SMI) crises. AI-assisted tools, including ChatUML, ChatGPT, Claude, and DeepSeek, were assessed for accuracy, clarity, efficiency, and cost. A sample SMI use case was used to compare six AI-generated UML diagrams against a human-created benchmark. Results show that AI tools can improve diagram creation efficiency, with ChatUML using Claude 3.5 Sonnet and DeepSeek Reasoning with R1 accessed through Perplexity performing best. The limitations in the other AI-generated outputs demonstrate the need for human oversight to ensure precision in healthcare applications. The findings suggest that the use of generative AI in system planning and design would accelerate the development of pathways for healthcare providers managing SMI crises, with the potential to extend to other care settings.

Pages: 30 to 36

Copyright: Copyright (c) IARIA, 2025

Publication date: May 18, 2025

Published in: conference

ISSN: 2308-4359

ISBN: 978-1-68558-270-8

Location: Nice, France

Dates: from May 18, 2025 to May 22, 2025