Recap - Innovatieve oplossingen voor chronisch zieken
BLOG: 128 BPM, or, the Semiotics of RECAP
Written by: Paul Brandt, Sander Stuijk, Twan Basten
Partner: Eindhoven University of Technology
Telehealth is considered a technological asset for the ageing society. It enables multidisciplinary teams to collaborate on improving the health of a patient without the need to be in the same physical location. This magnificent strength of telehealth at the same time represents its Achilles Heel. So many individuals collaborating, in a patient-specific, dynamic network of mutually influencing processes, results in an organizational complexity that requires facilitation by ICT. Putting all these ICT systems together is a major challenge that needs to be addressed before telehealth can deliver its promises.
Let’s take a relatively simple interoperability problem: several sensors (ECG, acceleration, skin conductance) being (re-)used by several different applications. A sensor delivers the same bits of data independent of the application. This data can have distinct purposes in different applications. Heartbeat can carry an indication of health in the care domain or an indication of performance potential in domains of sports. Yet, a heart rate reading of, say 128 BPM, in itself is insufficient to act upon and only when embedded in contextual awareness, such as the subject being at rest or running on a treadmill, it provides the necessary information to appropriately interpret the data.
Hence, before an application is able to appropriately interpret sensor data, it requires prior knowledge. Typically this knowledge is only available at design time, often remains implicit and tacit, and is always specific to each distinct sensor-enabled application. Consequently, applications implement their data interpretation implicitly within the application’s core functionality. This introduces a tight integration between the sensors and applications, revealing an IT characteristic that is considered the origin of the intolerable level of waste in healthcare spending (Fickenscher 2013): the inability to respond to changes, such as knowledge evolution or variations in situations, without costly software modifications.
We cannot afford to develop specific solutions for each combinatorial sensors/applications variation and need to cope with prior knowledge by improving on interoperability. Interoperability is defined as the ability of a system or product to work with other systems or products without special effort on the part of the customer*. From that definition we conclude that interoperability requires this prior knowledge about the sensor data to become (i) explicit, so that it can be addressed by the application; (ii) tangible, so that its relevance can be addressed by any application at runtime; and (iii) specified by open standards, in order to avoid vendor lock in and stimulate adoption (Brandt & Bruijning 2007).
Our first task is to make prior knowledge explicit: what do we need to specify in order to arrive at a meaningful communication between sensors and applications? This is where Semiotics, the study of signs and sign processes, comes into play since it makes a clear distinction between the syntactic, semantic and pragmatic aspects of communication. Sensors, mobile devices, eHealth records, decision support systems and social network services all need to become interoperable at all three aspects of communication.
During the last decade, much effort has been spent on achieving syntactic interoperability, resulting in widely adopted open communication protocols for accessing and exchanging sensor data. Less attention has been given to the interoperability of sensor networks and sensor applications at a semantic level. This hinders the reuse of sensors in different applications and the evolution of existing sensor networks and their applications. During the last year, we conducted research on an ontology-based approach and architecture to address this problem (Brandt et al. 2013). We introduced an ontology on context-aware examinology, ContoExam, and showed its role in the system architecture. We showed that by its application, open sensor networks emerge that can be loosely coupled with open sensor applications, enabling semantic interoperability in and across different sensor applications and even across different application domains. We are currently conducting research on some aspects of pragmatic interoperability and hope to report to you next year how improvements on this level of interoperability impact telehealth coordination!
*IEEE Standards Glossary, available at http://www.ieee.org/, accessed Sep 28th, 2012
Brandt, P. & Bruijning, J., 2007. XML for plug-and-play applications in service delivery. Gerontechnology, 6(3), pp.147-154.
Brandt, P. et al., 2013. Semantic interoperability in sensor applications. In 2013 IEEE Symposium Series on Computational Intelligence, SSCI 2013. Proceedings, Computational Intelligence in Healthcare and e-health, IEEE Symposium on (CICARE2013).
Fickenscher, K., 2013. President's column: interoperability--the 30% solution: from dialog and rhetoric to reality. Journal of the American Medical Informatics Association, 20(3), pp.593-594.