Novel Visualization of Large Health Related Data Sets
Abstract
Using retrospective data queries to understand what information clinicians seek from health care data, we have identified data elements and are looking at combinations of data elements used in queries. We are developing various visualization techniques that can be used to present the informational content in large databases, expecting that visualization of this data will present or "discover" information without specific hypotheses. Groups of related data elements are incorporated into visualizations that allow a quick comparison of data from a large population, with the ability to view trends over time within a chosen category. We are exploring the ability to compress petabytes of health care data representing many data elements into various groups of related data presented visually with an interface that allows the user to interactively explore the data elements to understand big data from the perspective of an entire population, different disease groups, ages, and other variables. There is the potential to detect causal relationships between various sets of data, which when applied to military EHR data may lead to improved health care and resiliency in military personnel, assist the DoD in strategic decisions related to personnel, and save millions of dollars in health care costs.
Document Details
- Document Type
- Technical Report
- Publication Date
- Mar 01, 2015
- Accession Number
- ADA624744
Entities
People
- David Borland
- Eugenia M. Heinz
- Igor Akushevich
- Vivian L. West
- William E. Hammond
Organizations
- Duke University