Reasoning with Contextual Data in Telehealth Applications

The notion of context-awareness is making its way in the area of mobile-health application designed to assist patients outside a hospital and medical professionals within a hospital. Such ubiquitous computing applications go beyond simple biomedical data-collection so the management of the context becomes more challenging, and the complexity of the context-based decision-algorithms warrants the use of full-blown inference engines. The context is a very volatile notion, so building a context-aware system means more than just creating a context ontology and defining inference rules for that model. This paper proposes mechanisms to dynamically manage the contextual decisions and actions. It devises a model and its properties for the process of taking each individual context-based decision. An architecture centered around this model is defined, integrating an off-the-shelf inference engine that reasons on context-data expressed in the OWL language. The architecture was refined and validated through the implementation of a prototype remote biomonitoring system. The prototype combines context-adaptation and awareness with the ubiquity of 3G network-coverage to allow for more personalized monitoring and better streamlining of information between the various health care players. It can be easily adapted and extended with extra inference-rules, additional actions, and more context items.