Research on the construction and application of pathology knowledge graph

Area of Science:

• Medical Education Technology
• Digital Pathology
• Knowledge Representation

Background:

• Pathology education struggles with fragmented knowledge, diagnostic accuracy, and clinical reasoning.
• Existing medical knowledge graphs lack multimodal integration and competency assessment.
• A Multimodal Knowledge Graph (MKG) with O-PIRTAS pedagogy was developed to address these limitations.

Purpose of the Study:

• To design and evaluate an integrated MKG with O-PIRTAS pedagogy for pathology education.
• To improve knowledge transfer, diagnostic accuracy, and clinical reasoning skills.
• To enable personalized competency assessment and development.

Main Methods:

• A pathology-specific MKG was constructed using Design Science Research Methodology.
• The MKG featured semantic modeling of diseases and cross-modal alignment of digital slides, animations, and clinical cases.
• Competency was quantified using embedded metrics (KII/MDA/CCAE) in a quasi-experiment involving 533 medical students.

Main Results:

• The MKG-O-PIRTAS group showed significantly higher exam scores (76.14 vs. 73.72) and an 86% lower misdiagnosis rate in high performers.
• Cognitive load was significantly reduced in the MKG-O-PIRTAS group (57.5 vs. 75.5).
• Over 80% of students reported enhanced knowledge integration and optimized learning processes.

Conclusions:

• The MKG-O-PIRTAS artifact effectively transforms pathology knowledge into clinical reasoning scaffolds.
• This approach facilitates personalized competency development in medical education.
• Future work will focus on adaptive scaffolding and integrating EMR modules for a replicable educational intelligence paradigm.