Related Experiment Videos
Architectural requirements for a multipurpose natural language processor in the clinical environment
C Friedman1, S B Johnson, B Forman
1Department of Computer Science, Queens College of CUNY, USA.
Proceedings. Symposium on Computer Applications in Medical Care
|January 1, 1995
Summary
This study presents a portable natural language processing system for automating clinical information encoding. The developed architecture enables broad clinical use across various applications, improving data extraction from medical reports.
Area of Science:
- Natural Language Processing
- Clinical Informatics
- Medical Data Management
Background:
- Automating clinical information encoding from text is challenging.
- Existing natural language processors are often limited to narrow domains.
- A need exists for widely applicable clinical information systems.
Purpose of the Study:
- To examine architectural requirements for portable natural language systems.
- To describe the architecture of a developed, versatile clinical information encoding system.
Main Methods:
- Investigated architectural requirements for system portability.
- Designed and developed a natural language processing system with a focus on broad applicability.
- Integrated the system into the clinical information system at Columbia-Presbyterian Medical Center.
Main Results:
- The developed system architecture supports portability across different domains.
- The system is capable of serving a variety of clinical applications.
- Routine encoding of clinical information from radiological reports is achieved.
Conclusions:
- The developed architecture facilitates the creation of widely usable clinical natural language processing systems.
- Portability is a key factor for successful integration and routine use in clinical environments.
- The system demonstrates successful routine encoding of clinical data from radiological reports.