Abstract

We describe the methodology and architecture of a knowledge-based, interactive visualization system that enables physicians and medical support personnel to draw conclusions from heterogeneous time-oriented clinical data. Our system employs domain-specific ontologies to produce temporal and statistical abstractions of data, and also as the basis for semantically-based browsing and visualization. This builds on previous work in data mining, temporal reasoning, and information visualization, but offers fundamental advantages over any isolated approach, by leveraging each off the others. We performed an evaluation of a prototype, leading us to conclude that users can indeed use the system to perform such semantically-based browsing in a reasonable amount of time.

Description

sdasda

Links and resources

Tags