Interactive Healthcare Big Data Analytics Platform under Simulated Performance

Operational Efficiencies and Simulated Performance of Big Data Analytics Platform over Billions of Patient Records of a Hospital System

“…Deploying our platform covered all the four of these spaces while being interactive was challenging. In previous study [51,54], we reported that Hadoop, HBase, and Apache Phoenix provided an excellent platform to perform SQL-like queries with high performance and accurate patient results. However, the ingestion of one billion records from Hadoop HFiles bulkloaded to HBase (indexed files) took weeks, while three billion took over two months on the infrastructure before finalizing the configurations.…”

“…These tools run interactively or via a batch job. Similar to Hadoop, all of this software is available through the module environments package [54,61]. Since the investigation was for user interactivity and usability the following was elaborated on:…”

“…Furthermore, the end user experience should include analytical tools with visualizations using web-based applications. To achieve this, a dynamic interactive BDA platform, and following our preliminary results from Chrimes, Moa, Zamani, and Kuo [54], established for healthcare application at the University of Victoria (UVic), in association with WestGrid, and Vancouver Island Health Authority (VIHA), Victoria, BC, Canada. Thus, our Healthcare BDA (HBDA) platform provided a proof-of-concept implementation and simulation of high performance using emulated patient data generated on WestGrid's Linux database clusters located at UVic.…”