Natural language generation of surgical procedures

Area of Science:

• Medical Informatics
• Natural Language Processing
• Clinical Terminology

Background:

• Surgical procedure rubrics are complex and require standardized conceptual representation.
• Existing systems lack robust methods for translating these representations into multilingual natural language.

Purpose of the Study:

• To present the compositional scheme developed by the GALEN-IN-USE project for surgical procedures.
• To demonstrate a natural language generation tool capable of translating these complex representations.

Main Methods:

• Developed a compositional scheme for conceptual representation of surgical operative procedure rubrics.
• Implemented a multilingual natural language generation tool.
• Adapted the generator using specific concept and relationship definitions.

Main Results:

• Successfully translated complex conceptual representations of surgical procedures into surface language.
• Demonstrated the adaptability of the natural language generation tool to the GALEN scheme's characteristics.

Conclusions:

• The GALEN-IN-USE compositional scheme provides a robust framework for representing surgical procedures.
• The developed natural language generation tool effectively bridges the gap between conceptual models and natural language for medical applications.