HistoriView: Implementation and Evaluation of a Novel Approach to Review a Patient Using a Scalable Space-Efficient...

Area of Science:

• Medical Informatics
• Data Visualization
• Human-Computer Interaction

Background:

• Patient data is often voluminous and intricate, leading to challenges in efficient review and analysis.
• Existing timeline visualizations struggle with merged event representations, causing recognition and interaction problems.
• Handling large patient datasets effectively remains a significant challenge in research.

Purpose of the Study:

• To develop a scalable and efficient timeline visualization tool, HistoriView, to enhance patient data review for research.
• To address the limitations of current timelines in handling complex and large-volume patient data.
• To improve event recognition, reduce access ambiguity, and facilitate efficient interaction with patient histories.

Main Methods:

• Developed HistoriView, a high-throughput, space-efficient timeline using nonstacking event representation.
• Implemented an overlay detection algorithm, y-shift visualization, and popup-based interaction for comprehensive analysis.
• Deployed an i2b2 HistoriView plugin utilizing split query and event reduction for efficient data delivery.

Main Results:

• HistoriView was found to be easy to use and learn, providing clear information without zooming.
• Participants preferred HistoriView over stacked timelines, citing satisfaction with display, ease of use, and efficiency.
• HistoriView efficiently visualized large datasets (32,630 records reduced to 2,019 artifacts) within acceptable timeframes, outperforming baseline limits.

Conclusions:

• HistoriView enables complete data exploration with a compact overview, beneficial for dense data and comparisons.
• The tool effectively handles large patient data, preserving original information efficiently.
• While effective, further improvements are suggested for exploring subconcept records.