21th AIAI 2025, 26 - 29 June 2025, Limassol, Cyprus

Incorporating Medical Assistants in eHealth Environments Using an Agentic RAG Approach

Kalathas Dimitrios, Menychtas Andreas, Maglogiannis Ilias, Tsanakas Panayiotis

Abstract:

  The evolution of Generative AI and AI Agents is continuously disrupting all fields of human-related activities, from daily life to enterprises and public sector. It is a common practice nowadays for all modern digital platforms to incorporate approaches which implement ChatBots using Large Language Models - LLMs to facilitate user interactivity, enrich the capabilities of the systems, and generally offer to the users an improved experience. Furthermore, in the field of healthcare, ChatBots and AI Agents are increasingly integrated into systems and applications for providing recommendations to the users, and interact with them to provide health and wellness-related information and respond to their questions. However, most of them use general-purpose LLMs, and therefore the respective results are not expected to be as accurate as required in the field of healthcare. Other approaches follow the method of fine-tuning to solve the limitation of inaccurate results but spend a lot of resources and time to achieve it. This work highlights the development of a novel Q&A ChatBot enhanced by innovative techniques to achieve high-accuracy results in the field of Medical Information. The advantage of the proposed prototype is that it integrates easily with healthcare platforms, both under development or already existing, to provide information and insights about medicines and patients' drug treatment.  

*** Title, author list and abstract as submitted during Camera-Ready version delivery. Small changes that may have occurred during processing by Springer may not appear in this window.