Related Experiment Videos
PDL: a definition language for trend pattern representation and detection in medicine
1Medical Computing laboratory, School of Computing, National University of Singapore, Singapore 11753. lij@comp.nus.edu.sg
Proceedings. AMIA Symposium
|February 5, 2002
Summary
This study introduces a Pattern Definition Language (PDL) for representing trend patterns in time-critical medical decision-making. PDL enhances pattern matching capabilities for complex temporal data analysis.
Area of Science:
- Computer Science
- Medical Informatics
- Data Analysis
Background:
- Medical decision-making in time-critical domains requires effective analysis of temporal data trends.
- Existing methods may lack the expressiveness to handle complex or irregular temporal patterns.
Purpose of the Study:
- To propose a novel Pattern Definition Language (PDL) for representing and manipulating trend patterns.
- To enhance the capabilities for analyzing temporal data in time-critical medical applications.
Main Methods:
- Development of a Pattern Definition Language (PDL), based on a modified Shape Definition Language (SDL).
- Extension of SDL's expressive power for temporal pattern matching.
- Incorporation of support for matching irregular length elementary patterns.
Main Results:
- PDL provides a robust framework for defining and manipulating trend patterns.
- The language effectively extends temporal domain expressiveness compared to SDL.
- PDL demonstrates applicability in time-critical medical scenarios.
Conclusions:
- PDL offers a powerful tool for medical decision support systems.
- The language facilitates more sophisticated analysis of temporal trends in healthcare.
- PDL's flexibility in pattern matching addresses limitations of previous approaches.