Presentation
Multi-Agent Systems (MAS) for Remote Healthcare with Enhanced Efficiency and Trust through Quantum-Model Methodology and Validation
SessionHC7: Modeling in Health Care
DescriptionArtificial intelligence (AI) chatbots have improved rapidly. However, these systems still face challenges in complex and time-sensitive issues where real-time awareness is imperative, such as remote emergent care. To address these limitations, a Multi-Agent System (MAS) was developed that employs a collection of AI agents with unique and distinct tasks, ranging from symptom analysis and user proficiency to risk assessment and information verification. In conjunction, these agents work together to enhance the clarity of output and thereby mitigate the hallucinatory effects associated with traditional single-agent systems. The trust dynamics of the human-AI team were measured quantitatively using a novel quantum model, implemented with Quiskit. In human subject experiments, the MAS system significantly reduced the number of follow-up questions and achieved higher trust scores than the single-agent system, indicating the model's validity. These results suggest that MAS-based systems can substantially improve the reliability and effectiveness of remote emergency care, offering a promising new direction for digital healthcare support. Future research will extend validation across broader populations and emergency scenarios.
Event Type
Industry/Practitioner Content
Lecture
TimeWednesday, October 15th12:10pm - 12:30pm CDT
LocationGrand C/D North
Health Care