Interactive Decision Support: A Framework to Improve Diagnostic Processes of Cancerous Diseases Using Bayesian Networks

Abstract

Recommendations for actions included in a Clin- ical Practice Guideline (CPG) provide a reference framework for medical experts during diagnostic processes. To support the implementation of these recommendations, we propose an interactive decision support. In order to realize this, the diagnostic processes in the CPGs of Mantle Cell Lymphoma (MCL) and Multiple Myeloma (MM) are formalized using activities of the Unified Modeling Language (UML). Based on UML activities, a Bayesian Network is generated. The resulting models enable an assistance function allowing for patient specific CPG recommendations and subsequently for a suitable as well as personalized diagnosis embedded in an interactive Decision Support System (DSS).