Development of a Rule-Based Knowledge Base for Digitalization of Standard Operating Procedures on Radiotherapy in Breast Cancer
Oncology · 2025
DOI: 10.1159/000549853 ↗Kurz gefasst
Fachartikel in der Zeitschrift Oncology, 2025. Verfasst von Fabio Dennstädt und 11 weiteren Autor:innen. Beteiligte Einrichtung: Department of Radiation Oncology, HOCH, Cantonal Hospital St.Gallen, St. Gallen, Switzerland, fabio.dennstaedt@insel.ch. Hauptthemen laut MEDLINE: Breast Neoplasms, Knowledge Bases. Die Arbeit nennt 25 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0).
Abstract (Originalsprache)
Introduction
Traditional text-based standard operating procedures (SOPs) have limitations in today’s data-rich environment, as they provide limited support for automated reasoning, contextual guidance, and real-time clinical decision-making. We developed a rule-based knowledge base to overcome these challenges through digitalization of SOPs on breast cancer radiotherapy.
Methods
We designed and developed a web application with a relational database structured around common data elements, a rule-based inference engine, and a responsive user interface (UI). The system’s information retrieval success was evaluated in a single-center study over 14 months, where nine radiation oncologists submitted clinical queries and provided feedback on the relevance of the results.
Results
The knowledge base incorporated 103 specific information entries covering 8 main topics, structured around critical clinicopathological features such as tumor stage, receptor status, grading, and lymphovascular invasion. During a 14-month testing period, nine radiation oncologists submitted 62 distinct clinical queries (e.g., determining boost indication for a patient of certain age and tumor situation). Of these queries, 56.5% were successfully answered as indicated by the user. Analysis of unsuccessful queries revealed that the information was mostly present but inaccessible due to user experience issues.
Conclusion
Our rule-based knowledge base demonstrates the feasibility of transforming static SOPs into interactive, evidence-linked resources. While promising, substantial user experience barriers remain. Future enhancements will prioritize UI improvements, sustainable knowledge-updating mechanisms, and AI integration to strengthen the system for practical use.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Department of Radiation Oncology, HOCH, Cantonal Hospital St.Gallen, St. Gallen, Switzerland, fabio.dennstaedt@insel.ch.
- Department of Radiation Oncology, Inselspital, Bern University Hospital, University of Bern, Bern, Switzerland, fabio.dennstaedt@insel.ch.
- TUM School of Computation, Information and Technology (CIT), Technical University of Munich, Heilbronn, Germany.
- Department of Radiation Oncology, Inselspital, Bern University Hospital, University of Bern, Bern, Switzerland.
- Department of Radiation Oncology, HOCH, Cantonal Hospital St.Gallen, St. Gallen, Switzerland.
Zitieren
APA
Dennstädt, F., Zink, J., Cihoric, N., Weder, D., Glatzer, M., Crommelinck, J., Süveg, K., Fischer, G. F., Hà, H. L., Brügge, D., Plassilm, L., & Putora, P. M. (2025). Development of a Rule-Based Knowledge Base for Digitalization of Standard Operating Procedures on Radiotherapy in Breast Cancer. Oncology. https://doi.org/10.1159/000549853
Vancouver
Dennstädt F, Zink J, Cihoric N, Weder D, Glatzer M, Crommelinck J, et al. Development of a Rule-Based Knowledge Base for Digitalization of Standard Operating Procedures on Radiotherapy in Breast Cancer. Oncology. 2025. doi:10.1159/000549853
BibTeX
@article{Dennstdt2025,
author = {Dennstädt, Fabio and Zink, Johannes and Cihoric, Nikola and Weder, Dagmar and Glatzer, Markus and Crommelinck, Julie and Süveg, Krisztian and Fischer, Galina Farina and Hà, Hông-Linh and Brügge, Detlef and Plassilm, Ludwig and Putora, Paul Martin},
title = {Development of a Rule-Based Knowledge Base for Digitalization of Standard Operating Procedures on Radiotherapy in Breast Cancer},
journal = {Oncology},
year = {2025},
doi = {10.1159/000549853}
}
EUVIMED