Creating a flexible integration framework for physiological modeling

Integration Science is the study of the processes and phenomena required to build effective combined models from disparate component or partial models. Integration is usually treated very differently in different subject domains, but there are several commonalities across all of them: the explicit identification and processing of assumptions, collections of interface elements that convert as much as possible between different sets of assumptions, and monitors that track the validity of assumptions during operation. Our research results in Integration Science include "Wrappings", the Knowledge-Based integration infrastructure that we developed to assist us with this scale of integration problem. Our main innovation is our focus on making the processes of an integration problem as well as the pieces all explicit and selectable, so that they can be studied and possibly altered. This paper is about the application of our Wrapping integration infrastructure to combining physiological models of human beings, with the aim of producing effective and reliable models of the physiological health of individuals, so that indications and warnings about their physiological state can be tracked, and also anticipated. The application will be to deep space missions, in which the patients cannot be reached in any useful amount of time.