# Interactive Information Visualization to Explore and Query Electronic Health Records
Interactive Information Visualization to Explore and Query Electronic Health Records
This paper reviews the state-of-the-art of information visualization systems for exploring and querying EHRs, by evaluating their techniques, designs and effectiveness in knowledge discovery. The authors intend to identify potential research directions for researchers and designers of Information visualization to support EHR systems.
The authors classify 14 information visualization systems of EHR in a hybrid of categorical and hierarchical classification.
The authors focus on 14 information visualization systems of EHR. Multi-Patient record support is the ability to analyze a single patient or a cohort of patients, the authors conclude that this has the largest impact on the system.
In the context of single record support, four critical features are identified as key features:
- Events over Time visualize a series of events over time with their categories.
- Numerical Data over Time depicts multiple numerical variables over time.
- Heterogeneous Data on a Common Time Axis is the combination of events and numerical data visualized using different visual designs and techniques.
- Snapshots of Patient State, which heavily relies on the use of glyphs and animations to encode multiple data variables.
The authors acknowledge that multiple record support is required to support quality control and clinical research. Compared to a single record, it provides less detailed visualizations of a single patient but is able to identify clusters and patterns within a cohort of patients. This topic is subdivided into four categories:
- Event Sequences which are similar to Events over Time but with multiple record support. However fewer details are provided as the trade-off.
- Expressive Temporal Queries are used for querying records with user-defined queries such as time intervals or event sequences (e.g. heart attack followed by stroke).
- Multiple Patient Records on a Common Time Axis.
- Snapshots of Multivariate Patterns in Patient Cohorts, this is achieved through the use of animations to find clusters in the cohort.
The authors also classified 14 EHR visualization systems by their interactive features [See Table 5.3].
They also briefly reviewed 32 other EHR systems in conjunction with 6 related surveys.
- Different user tasks require different visualization techniques. In most cases, visualizations need to be tailor-made for each diagnosis.
- Increase amount and complexity in data, this requires not only multiple visualization techniques but also the combination of automated analysis techniques and visualizations.
- Patients keep their own data and are interested in visualising it, however, existing tools are either not open for individual use (patients personal use) or too complex that they overwhelm casual users.
- The transition from multiple patient analysis to single patient analysis is not well studied.
- Most systems are missing advanced interactive visualization features to support clinical tasks.
- A Large amount of data increases the difficulty in data management and knowledge discovery.
- Systematic errors in data are hard to identify and rectify.
Information Visualization of EHR.
Rind, A., Wang, T. D., Aigner, W., Miksch, S., Wongsuphasawat, K., Plaisant, C., & Shneiderman, B. (2011). Interactive Information Visualization to Explore and Query Electronic Health Records. Foundations and Trends® in Human–Computer Interaction, 5(3), 207–298. https://doi.org/10.1561/1100000039 ↩︎