Towards a Multi-Agent System for Medical Records Processing and Knowledge Discovery

Abstract

Medical files and observation papers have always been an important source of knowledge. Unfortunately, most of the times they are still stored as physical documents either printed or handwritten, thus making it difficult to transfer this precious information from one place to another, or centralizing and extracting new knowledge from it. But with nowadays advancements in computer science this problem may be handled. In this paper we propose a patient centered Multi-agent system project that is able to extract relevant information from the patients' health records and store that knowledge on a centralized data-store, based on a predefined ontology scheme. The system's purpose is to standardize and enrich the knowledge by performing various mining tasks on the given text. The final purpose is to provide different hospital departments with a tool which they can query for useful information about the patients' medication, treatments and/or scheduled tests and exams and get suggestions regarding their treatment intentions.