Interactive Visualization Methods to Support Protocol-Based Care

Team

Wolfgang Aigner, Vienna University of Technology, Institute of Software Technology and Interactive System, Vienna, Austria
Silvia Miksch, Vienna University of Technology, Institute of Software Technology and Interactive System, Vienna, Austria

Contact Person

Wolfgang Aigner

Project

The aim of this project is the development of interactive visualization methods to support protocol-based care.

We provide multiple simultaneous views to cover different aspects of a complex underlying data structure of treatment plans and patient data. The tightly coupled views use visualization methods well-known to domain experts and are designed to facilitate users' tasks. CareVis enables a meaningful navigation, provides annotations on demand for not overwhelming the viewer, and helps to keep orientation by using focus+context techniques, thus increasing the flexibility in working with treatment plans and patient data. Moreover, we have examined the usefulness of our approach performing a 3-step evaluation process including user study, design reviews, and prototype evaluation.

Views
Basically, the underlying data structure is divided along the lines of logical structure and temporal aspects. Hence, we provide a Logical View and a Temporal View along with a QuickView Panel. These distinct views are presented simultaneously and divide the screen in the following manner (see image below). The QuickView Panel is located on top of the screen displaying the most important patient parameters and plan variables at a prominent position. Below that, the screen is divided vertically by the logical view on the left-hand side and the temporal view on the right-hand side. The logical view presents treatment plans in terms of their logical structure (hierarchical decomposition, plan elements, execution order, conditions). The temporal view on the other side focuses on the temporal aspects of treatment plans and measured patient data as well as plan variables (temporal aspects of plan elements, temporal uncertainties, hierarchical decomposition).

Logical View
The logical view on the left-hand side of the screen provides a representation of the treatment plan specification data. The applied visualization technique AsbruFlow is based on the idea of flow-chart-like clinical algorithm maps that are well known amongst physicians. This concept has been extended in order to be able to depict the characteristics of a treatment plan formulated in the plan specification language Asbru.

Temporal View
The temporal representation of treatment plans is based on the idea of LifeLines. This concept has been extended for enabling the display of hierarchical decomposition as well as the complex time annotations used in Asbru. These new visual elements are called LifeLines+ and PlanningLines, respectively. LifeLines+ allow the interactive representation of temporal intervals with hierarchical decomposition and simple element characteristics. On top of that, PlanningLines allow the depiction of temporal uncertainties via a glyph consisting of two encapsulated bars, representing minimum and maximum duration, that are bounded by two caps that represent the start and end intervals. Encapsulated bars that can be shifted within the constraints of two mounted caps resemble the glyph’s mental model.

QuickView Panel
A separate possibility to display currently valid variable and parameter values is the so-called QuickView Panel in the top part of the application window (see image below). The panel consists of rectangular areas that can be assigned to the available parameters and variables. A single item shows the current value along with its name, unit, and a trend indicator. Thus, the QuickView Panel allows to monitor the most important values by putting them at a prominent position, enlarged in size and without the need for displaying the complete history in an additional facet.

Logical View
Temporal View
QuickView Panel

Asbru plans
+
+

    Time-oriented data

+

    Logical sequences
+

    Hierarchical decomposition
+
+

    Non-uniform element types
+
◦

    Conditions
+

Parameters and variables

+
+

Data characteristics in views
entirely represented ( + ), partly or implicitely represented ( ◦ ), or not represented (empty).

Papers

W. Aigner, S. Miksch: CareVis: Integrated Visualization of Computerized Protocols and Temporal Patient Data; Artifical Intelligence in Medicine (AIIM), Vol. 37, No. 3, p. 203-218, Elsevier, July, 2006.

W. Aigner, S. Miksch: CareVis: Integrated Visualization of Computerized Protocols and Temporal Patient Data;
Presentation accepted for: Workshop on Intelligent Data Analyis in Medicine and Pharmacology (IDAMAP-2004), Stanford, USA; 06-09-2004; in: "Workshop Notes of the Workshop on Intelligent Data Analyis in Medicine and Pharmacology", (2004).

W. Aigner, S. Miksch: Supporting Protocol-Based Care in Medicine via Multiple Coordinated Views; in: Proceedings International Conference on Coordinated and Multiple Views in Exploratory Visualization (CMV 2004), IEEE, (2004).

W. Aigner, S. Miksch: Communicating the Logic of a Treatment Plan Formulated in Asbru to Domain Experts; in: Computer-based Support for Clinical Guidelines and Protocols, K. Kaiser, S. Miksch, S. Tu (ed.); IOS Press, 101 Studies in Health Technology and Informatics (2004), ISBN 1 58603 412 X; 1 - 15.

W. Aigner: Interactive Visualization of Time-Oriented Treatment Plans and Patient Data; Master's thesis, Supervisor: S. Miksch; Institut für Softwaretechnik und Interaktive Systeme, 2003.

Images

To retrieve an enlarged version of the images, click on them:

CareVis application window

AsbruFlow example

LifeLine+, PlanningLines

AsbruFlow structure

Downloads

	>> Download CareVis research prototype
	Filename: execview_prototype.zip Needed application: Java JRE 1.4.2	Size: 944 kB Time to download: approx. 2:15 min (56k Modem)

	>> Download PlanningLine Viewer prototype (zTimeView)
	Filename: zTimeView.zip Needed application: Java JRE 1.4.2	Size: 2 MB Time to download: approx. 5 min (56k Modem)

Related Work

Zoomable TimeView (Demo of a zoomable version of "PlanningLines")
LifeLines for Visualizing Patient Records (C. Plaisant et al. - University of Maryland)

Funding

This project is supported by “Fonds zur Förderung der wissenschaftlichen Forschung - FWF” (Austrian Science Fund), grant P15467-INF.