Digital Twin for Mental Health: Designing a Multi-Layer Digital Twin for Mental Health with Explainable AI

Sammanfattning

Introduction: The increasing complexity of mental healthcare and the growing demand for proactive and personalized treatment approaches have created a need for technologies that support continuous monitoring, forecasting, and clinical decision-making. Although Digital Twins have gained increasing attention in healthcare research, previous studies have highlighted challenges related to conceptual modeling and explainability of artificial intelligence systems. Research Question: The primary research question of this study is: “How can a multi-layered Digital Twin for mental healthcare be conceptually modeled to incorporate forecasting and Explainable Artificial Intelligence?” Method: This study adopts a Design Science Research methodology. The conceptual Digital Twin artifact was developed through an exploratory literature search, iterative modeling, and expert evaluations involving both a clinical expert and a Digital Twin expert. Results: The results indicate that the proposed conceptual model incorporates key Digital Twin characteristics, including bidirectional data flows, continuous monitoring, validation, forecasting capabilities and explainability. The findings also emphasize the significance of patient participation and human oversight in Artificial Intelligence supported mental healthcare systems. Discussion: This study contributes to existing research by presenting a conceptual model for Artificial Intelligence based Digital Twins in mental healthcare using a conventional modeling language. However, the study is limited by its conceptual scope and lack of technical implementation. Future research should address challenges related to practical implementation, patient safety and data security.