Related Experiment Videos
Toward a medical-concept representation language. The Canon Group
D A Evans1, J J Cimino, W R Hersh
1Department of Philosophy, Carnegie Mellon University, Pittsburgh, PA 15213-3890, USA.
Summary
Researchers are developing a new medical language representation to improve data exchange and application development. This deeper formalism aims to make implicit relations visible and resolve ambiguities in medical texts.
Area of Science:
- Medical Informatics
- Computational Linguistics
- Knowledge Representation
Background:
- Current medical data representation is often inadequate, limited, or rigid.
- Existing methods are idiosyncratic, hindering generalization and extension across applications.
- There is a need for a more robust formalism for medical data.
Purpose of the Study:
- To design a general schema for medical language representation.
- To develop resources and procedures for mapping medical language into a deeper representation.
- To make implicit relations explicit and resolve ambiguities in medical data.
Main Methods:
- Examining large volumes of medical texts from various sub-domains.
- Identifying candidate medical concepts within these texts.
- Developing general rules and representations for attributes and values for uniform concept expression.
Main Results:
- Identification of candidate concepts for a general medical language schema.
- Development of foundational rules for uniform expression of medical concepts.
- Progress towards a formalism that reveals hidden attributes and resolves vague references.
Conclusions:
- A deeper representation formalism is crucial for advancing medical informatics.
- Standardizing medical language representation will enhance data exchange and application development.
- The proposed schema aims to overcome limitations of current medical data characterization.