C. Urach, H. Piringer, N. Popper, F. Breitenecker:

"Parallel sets as interfactive visualization approach ior analyzing markov models";

Value In Health,15(2012), 7; 473 - 474.

OBJECTIVES: Commonmethods to visualize Markov states over time (e.g., Markovcycle

trees or state probability graphs) do not scale well to many cycles and have

limitations concerning the perception of proportions. An objective of the IFEDH

research project (FFG grant number 827347) was to overcome these limits by investigating

new visualization methods of Markov models and their results. METHODS:

Inspired by the "Parallel Coordinates", an interactive technique called Parallel Sets

has been developed for visualizing multidimensional categorical data. The visualization

lays out axes in a parallel way where each axis represents one categorical

dimension. Within each axis, boxes represent the categories which are scaled according

to the respective frequencies. Applied to Markov Models, the categorical

dimensions correspond to the various cycles. Joint probabilities of categories from

adjacent axes are shown as parallelograms connecting the respective categories.

The parallelograms can be interpreted as the number of patients transiting from one state to another. Depending on the purpose, the color of the parallelograms

indicates the categories of a chosen cycle or could refer to additional attributes of

the patients like age or sex. RESULTS: State probability and survival curves merely

show specific aggregates of the data while classic Markov trace visualizations with

for example bubble diagrams do not visualize data in a sense that would facilitate

a detection of proportions and trends. Applying Parallel Sets to analyze Markov

models provides an interactive visualization technique where changing the reference

Markov cycle is as easy as highlighting particular dimensions, thus enabling

the exploration of the progress of patient cohorts with certain characteristics

through the model. CONCLUSIONS: Model development always requires thorough

analysis of its structure, behavior and results. Parallel Sets enable an intuitive and

efficient visualization technique for presentation purposes as well as exploratory

analysis.

Created from the Publication Database of the Vienna University of Technology.