VisuExplore – Gaining New Medical Insights from Visual Exploration

Submitted by alex on Fri, 10/05/2012 - 19:19
Aim

The project's objective is the implementation of a flexible, interactive visualization environment for time-oriented, medical data and information. These methods and tools will facilitate data analysis and interpretation tasks for physicians and clinical personnel in order to support medical treatment and enhance quality of care.

Duration

March 2008 – March 2011

Funding

This project was funded by the BRIDGE program of the Austrian Research Promotion Agency (FFG) (Project number: 814316) and carried out by a high-class consortium consisting of Danube University Krems, Vienna University of Technology, NÖGUS, NÖ Landeskliniken-Holding, Landesklinikum Krems, and systema Human Information Systems.

Contact
Status
finished

VisuExploreThe application of modern technology in clinical practice leads to a massive increase in quantity and complexity of electronically acquired medical data and information. Current commercial clinical information systems are faced with the explosion of available medical data and users of these systems ask for alternative, more intuitive solutions.

Interactive visual representations can be used to avoid information overload and to make information from multiple heterogeneous data sources more comprehensible. Especially in the medical domain time-oriented data plays a central role. These representations encompass changes of patients' condition, therapeutic actions, interventions, and measures as well as trends, patterns, and relationships between these variables over time.

Central to the visualization of medical data is not the presentation or graphical rendering but the gaining of new insights and knowledge from intensive active examination of the collected data and information. Static visualizations do not support such an analytical process because they allow for passive viewing only. Visual exploration requires flexible methods for interaction and manipulation.

The project VisuExplore had been initiated by Silvia Miksch and aimed at the development of such methods and tools. The project's objective was to provide flexible and interactive visualization environments for time-oriented, medical data and information. These methods and tools should facilitate data analysis and interpretation tasks for physicians and clinical personnel in order to support medical treatment and enhance quality of care. Throughout the project, we developed the two research prototypes VisuExplore and TimeRider for two different application scenarios:

VisuExplore is a flexible interaction and visualization framework for an electronic health record of a single patient. It presents the development of multiple heterogeneous medical variables over time. A particular focus was set on a high level of usability by using simple and easy-to-understand visual representation methods. In addition, VisuExplore is an extensible platform for further research developments. For example, three further visual representation methods were implemented and integrated during the project: a semantic zoom chart (based on [Bade et al., 2004]) to combine quantitative data (e.g., body temperature) and qualitative abstraction thereof (e.g., “high fever”), a horizon graph (based on [Reijner, 2008]) for a particularly space-efficient display of large amounts of numerical variables, and a document browser.

TimeRider uses and extends animated scatter plots for the analysis of patient cohorts over time. Thereby, we studied the use of animation, which is an important visualization alternative to time axes. Furthermore, we had to tackle additional challenges in order to explore patient cohorts effectively.

We succeeded in cooperating with physicians and involving them in the design and development process. Thus, we could base the interaction and visual representation methods on medical requirements and evaluate them in practice. This is an important prerequisite for usable and useful outcomes, but is often not achieved when facing the workload of clinicians.

In the evaluation of VisuExplore, we accumulated important scientific results, which are not only relevant for our prototype but also for comparable interaction and visual representation methods. For example, our subjects preferred familiar visual representations to more innovative methods. Furthermore, they applied largely diverging strategies to gain insights. This implies that a visual interface with a large repertoire of alternative interaction methods serves them better. In addition to usability issues, we gained important insights about users' cognitive strategies.

Related Work:

Publications

2013
Alexander Rind, Taowei Wang, Wolfgang Aigner, S. Miksch, Krist Wongsuphasawat, Catherine Plaisant, Ben Shneiderman, "Interactive Information Visualization to Explore and Query Electronic Health Records", Foundations and Trends in Human-Computer Interaction, vol. 5, pp. 207-298, 2013. paper
2012
Margit Pohl, Silvia Wiltner, S. Miksch, Wolfgang Aigner, Alexander Rind, "Analysing Interactivity in Information Visualisation", KI - Künstliche Intelligenz, vol. 26, pp. 151-159, 2012. paper
David Riano, Annette Teije, S. Miksch, "Knowledge Representation for Health-Care: AIME 2011 Workshop KR4HC 2011: Revised Selected Papers", Lecture Notes in Artificial Intelligence, vol. 6924, pp. 171, 2012. paper
2011
Alexander Rind, Wolfgang Aigner, S. Miksch, Silvia Wiltner, Margit Pohl, Felix Drexler, Barbara Neubauer, Nikolaus Suchy, "Visually Exploring Multivariate Trends in Patient Cohorts Using Animated Scatter Plots", Ergonomics and Health Aspects of Work with Computers, Proceedings of the International Conference held as part of HCI International 2011, pp. 139–148, 2011. paper
Paolo Federico, Wolfgang Aigner, S. Miksch, Florian Windhager, Lukas Zenk, "A Case Study on Visual Analytics of Dynamic Networks", The Third International UKVAC Workshop on Visual Analytics (VAW 2011), 2011. paper
Wolfgang Aigner, "Understanding the Role and Value of Interaction: First Steps", Proc. of the International Workshop on Visual Analytics (EuroVA 2011), pp. 17--20, 2011.
Paolo Federico, Wolfgang Aigner, S. Miksch, Florian Windhager, Lukas Zenk, "A Visual Analytics Approach to Dynamic Social Networks", Proceedings of the 11th International Conference on Knowledge Management and Knowledge Technologies (i-KNOW), Special Track on Theory and Applications of Visual Analytics (TAVA), pp. 47:1–47:8, 2011. paper
Alexander Rind, Wolfgang Aigner, S. Miksch, Silvia Wiltner, Margit Pohl, Thomas Turic, Felix Drexler, "Visual Exploration of Time-oriented Patient Data for Chronic Diseases: Design Study and Evaluation", Proceedings of USAB 2011: Information Quality in e-Health, pp. 301–320, 2011. paper
2010
Alexander Rind, S. Miksch, Wolfgang Aigner, Thomas Turic, Margit Pohl, "VisuExplore: Gaining New Medical Insights from Visual Exploration", Proceedings of the 1st International Workshop on Interactive Systems in Healthcare (WISH@CHI2010), pp. 149-152, 2010. paper
Alexander Rind, Wolfgang Aigner, Thomas Turic, S. Miksch, Margit Pohl, "VisuExplore: Gewinnung neuer medizinischer Erkenntnisse durch visuelle Exploration", Tagungsband der eHealth2010: Health Informatics meets eHealth, pp. 273-278, 2010. paper
2009
Wolfgang Aigner, Alessio Bertone, Tim Lammarsch, S. Miksch, Alexander Rind, "Interactively Exploring Time-Oriented Data", Workshop Notes of CHI 2009 workshop "Interacting with temporal data" at Conference on Human Factors in Computing Systems (CHI 2009), 2009. paper
Alexander Rind, "Interactive Information Visualization in Patient Care and Clinical Research: State of the Art", , 2009.
2008
Alexander Rind, "Datenflut im medizinischen Alltag - PatientInnendaten im Überblick", TIMNEWSupgrade, vol. 42, pp. 20–21, 2008. paper
Wolfgang Aigner, "VisuExplore - Gewinnung neuer Medizinischer Erkenntnisse durch visuelle Exploration", TIMNEWSupgrade, vol. 41, pp. 24, 2008. paper