Towards a 'Big' Health Data Analytics Platform

Health is generating large volumes of data that can provide invaluable insights into clinical and operational aspects of healthcare delivery. There is a general lack of specialized and integrated health data analytics platforms that offer technical methods to support the entire health data analysis pipeline -- i.e. health data selection, integration, analysis, visualization and sharing. This paper proposes the technical architecture of a health data analytics platform that offers a technical solution for analyzing 'big' health data originating from multiple sources with heterogeneous terminologies and schemas. A key aspect of the architecture is data standardization, where we have used SNOMED-CT as a terminology standard to standardize health data from multiple sources. We offer a single step health data integration solution where users can select the data sources and the data elements from multiple sources, and our platform performs the data standardization and data integration to prepare an integrated dataset. We present a case study involving large volumes of laboratory data that is integrated and analyzed using our platform.