This chapter focuses on the central role of information visualization in health analytics. From the early x-rays to 3D volume visualizations, rapid progress has been made. However, the most exciting growth is now in the area of information visualization, which offers interactive environments and analytic processes that help support exploration of electronic health record (EHR) data, monitoring, or insight discovery. For example, a health organization might want to investigate patterns of drug prescriptions in patients with asthma and compare prescribing practices with current guidelines. Temporal patterns are critical to this analysis, and interactive visualizations are beginning to support powerful temporal queries, present rich result summaries, and offer fluid interactions to identify the clinically relevant patterns hidden in the data. Visualization should soon help clinicians identify cohorts of patients who match selection criteria for clinical trials or need to be brought back to the office. Visualization can also reveal data quality problems, which are common when repurposing clinical data for secondary analysis. After a quick summary of the state-of-the-art of information visualization systems for exploring and querying EHR data, we describe in detail one recent system (EventFlow) developed by the Human-Computer Interaction Lab at the University of Maryland and illustrate its use with an asthma prescription study example.
分享
引用
