Cardiology eHealth messages routing policies management driven by Dynamic Bayesian Networks

The emergence of eHealth and the proliferation of mobile healthcare computing devices has led to a large increase in message transfers among remote healthcare providers and patients. Different scenarios in cardiology, such as the follow up of chronic heart diseases at home obviously require intelligent and reliable eHealth messages communication policies to proactively react in case of unexpected events (exceeded deadlines for reply,...) or context changes (chest pain increase,...). We propose a cardiology eHealth message modeling process that represents an orchestration of information systems and services for the support of context-aware, personalized, intelligent and adaptive routing policies. Several contextual data from the source (patient clinical signs), the target (healthcare professional localization), and the message content itself are taken into account for processing the message transfers. The message content is compliant with the HL7 Reference Information Model specifications. We finally demonstrate the process of inferring routing parameters such as the requested healthcare professional profile type and the routing means in function of different context values by means of Dynamic Bayesian Networks, and we highlight the routing policy adaptation process.