THEMPO: a knowledge-based system for therapy planning in pediatric oncology

Area of Science:

• Medical Informatics
• Artificial Intelligence in Medicine
• Oncology

Background:

• Pediatric oncology requires complex, protocol-directed treatment planning.
• Managing diverse patient data and treatment protocols is challenging.
• Existing systems may lack integrated knowledge representation for therapy management.

Purpose of the Study:

• To describe the THEMPO (Therapy Management in Pediatric Oncology) system.
• To present a knowledge-based approach for protocol-directed therapy planning.
• To detail the system's architecture for representing oncological knowledge and patient data.

Main Methods:

• Development of a knowledge-based system (THEMPO) utilizing semantic networks and graph grammars.
• Implementation of knowledge acquisition tools for domain authoring.
• Integration of medical problem solvers for reasoning about therapeutic and diagnostic timetables.
• Creation of a patient record system based on semantic networks.

Main Results:

• THEMPO effectively supports protocol-directed therapy planning and configuration in pediatric oncology.
• The system employs a semantic network and graph grammars for comprehensive knowledge representation.
• Medical problem solvers can reason about patient-specific treatment and diagnostic schedules.
• A patient record system accurately represents the course of oncological therapy.

Conclusions:

• Knowledge-based systems like THEMPO can enhance the precision of pediatric oncology treatment planning.
• The use of semantic networks and graph grammars provides a robust framework for medical knowledge representation.
• THEMPO offers a valuable tool for managing complex therapeutic and diagnostic pathways in pediatric cancer care.