SAPPHIRE: A Service-oriented Architecture for a Personalized Open-Source Health Information Research Exchange.

摘要

Lack of ability to translate biomedical information into meaningful knowledge for discovery remains to be the main challenge. There is a need for an innovative approach to manage disparate patient-level data in order to obtain scientific evidence for personalized treatment with higher efficacy and lower toxicity. Therefore, standardized data should be conceptually described in a logical way and processed in a compatible manner to address relevant research questions. The central argument of this dissertation is how to simplify data complexity using Service-Oriented Architecture (SOA) collaborative platform that has the ability to embrace a personalized research design and run complex operations. SAPPHIRE is the acronym for Service-oriented Architecture for a Personalized Open-Source Health Information Research Exchange. The architecture (also referred to as 'application platform') is built on open-source standards and open tool integration platform as a candidate system for biomedical knowledge discovery. Relevant research questions were addressed including, (1) what is the right SOA design? (2) what are the minimum data elements and standards that should be contained in the data models? (3) what is the best method to aggregate, integrate and exchange metadata? The hypotheses were formulated in the context of high-level data connectivity, which relate the preferred architectural design to its capability. Each hypothesis was tested against the type of connectivity, number of applications, model type, number of logical bridges and cost effectiveness. Technical architectural design specifications and requirements analysis for metadata integration was conducted. New approach to solve the non-systematic method to aggregate and process different data types from disparate sources was presented. To evaluate the architecture prototype, a proof-of-concept on a set of models was conducted. New data models were developed as an integral part of the newly designed architecture to generate a formal and valid set of results. This research attempts to develop an open-source integrative platform that improves the continuum of knowledge-driven research, ensures cost-effective access to data resources and enables freedom of innovation. In short, SAPPHIRE is an application framework that powers multiple open-source applications to integrate relevant "typic" metadata sets into contextual multi-dimensional models for data mining to accelerate knowledge discovery in biomedicine.;Keywords: architecture, service-oriented, personalized, open-source, standards, integration, biomedicine