Guideline-based decision support for the mobile patient incorporating data streams from a body sensor network

We present a mobile decision support system (mDSS) which helps patients adhere to best clinical practice by providing pervasive and evidence-based health guidance via their smartphones. Similar to some existing clinical DSSs, the mDSS is designed to execute clinical guidelines, but it operates on streaming data from, e.g., body sensor networks instead of persistent data from clinical databases. Therefore, we adapt the typical guideline-based architecture by basing the mDSS design on existing data stream management systems (DSMSs); during operation, the mDSS instantiates from the guideline knowledge a network of concurrent streaming processes, avoiding the resource implications of traditional database approaches for processing patient data which may arrive at high frequencies via multiple channels. However, unlike typical DSMSs, we distinguish four types of streaming processes to reflect the full disease management process: Monitoring, Analysis, Decision and Effectuation. A prototype of the mDSS has been developed and demonstrated on an Android smartphone.