These include: (i) the translation of services into rules enabling a rule-based system, which triggers events based on new data; (ii) the construction of a service workflow in order to gather all the necessary information in notifying the physician of the patients state; and (iii) the mapping of the service workflow to the available devices for execution depending on the location of the physician and their quality attributes, such as location, screen size and on-status. In order to achieve such functionality, the following devices and technologies are available in our paradigm, detailed in Section 6.1:Wireless medical devices and sensorsacquiring patient’s vital signs.An Indoor Localization System(ILS) consisting of a network of sensors used as anchor points in order to pinpoint the location of a person.

A Monitoring application [5] recording the aforementioned patient bio signals and hosting risk assessment algorithms to enable the alerting process.An ontology-drivenapplication intelligence capable of reasoning on the patient’s congestive heart failure (CHF) profile and the available devices for monitoring and notification.The remainder of the paper is structured as follows: Section 2 provides an overview of existing frameworks supporting the real-time monitoring of patients. In order to illustrate the AmI framework, Section 3 introduces a general clinical scenario related to monitoring patients diagnosed with congestive heart failure (CHF).

Subsequently, Section 4 elaborates on the implementation details of the architectural components of the AmI followed by a description of the developed ontologies for the formal definition of the medical patient profiles and the specification of device characteristics in Section 5. Section 6 describes the validation of the clinical scenario, including the supportive technologies specific to our implementation. A discussion on the benefits of a semantically-enhanced remote monitoring AmI is presented in Section 7. Finally, Section 8 ends with concluding remarks and future improvements.2.?Related WorkAdvanced decision support systems are in use at a handful of academic medical centers [6�C8]. Their predefined rules generate reminders to physicians based on clinical data, such as laboratory results, visit diagnosis, coded medications prescribed in the clinic and vital signs collected on encounter Drug_discovery forms.

Software alerts track patients�� vital trends and intervene on time before complications occur. Such facilities linked via telemedicine and computer monitors to the hospital rooms provide for the required around-the-clock specialized care of hundreds of patients. Although this approach emphasizes direct physician interaction and extensive clinical decision support, it lacks support for complex scenarios capturing multiple chronic conditions of a patient.