Secondary Use of Clinical Problem List Entries for Neural Network-Based Disease Code Assignment
Studies in Health Technology and Informatics · 2023
DOI: 10.3233/shti230267 ↗Kurz gefasst
Fachartikel in der Zeitschrift Studies in Health Technology and Informatics, 2023. Verfasst von Markus Kreuzthaler und 3 weiteren Autor:innen. Beteiligte Einrichtung: Institute for Medical Informatics, Statistics and Documentation, Medical University of Graz, Austria. Hauptthemen laut MEDLINE: Neural Networks, Computer, Language. Die Arbeit wurde bisher 1-mal zitiert (Crossref). Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0).
Abstract (Originalsprache)
Clinical information systems have become large repositories for semi-structured and partly annotated electronic health record data, which have reached a critical mass that makes them interesting for supervised data-driven neural network approaches. We explored automated coding of 50 character long clinical problem list entries using the International Classification of Diseases (ICD-10) and evaluated three different types of network architectures on the top 100 ICD-10 three-digit codes. A fastText baseline reached a macro-averaged F1-score of 0.83, followed by a character-level LSTM with a macro-averaged F1-score of 0.84. The top performing approach used a downstreamed RoBERTa model with a custom language model, yielding a macro-averaged F1-score of 0.88. A neural network activation analysis together with an investigation of the false positives and false negatives unveiled inconsistent manual coding as a main limiting factor.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Institute for Medical Informatics, Statistics and Documentation, Medical University of Graz, Austria.
- Department of Information and Process Management, Steiermärkische Krankenanstaltengesellschaft m.b.H. (KAGes), Graz, Austria.
Zitieren
APA
Kreuzthaler, M., Pfeifer, B., Kramer, D., & Schulz, S. (2023). Secondary Use of Clinical Problem List Entries for Neural Network-Based Disease Code Assignment. Studies in Health Technology and Informatics. https://doi.org/10.3233/shti230267
Vancouver
Kreuzthaler M, Pfeifer B, Kramer D, Schulz S. Secondary Use of Clinical Problem List Entries for Neural Network-Based Disease Code Assignment. Studies in Health Technology and Informatics. 2023. doi:10.3233/shti230267
BibTeX
@article{Kreuzthaler2023,
author = {Kreuzthaler, Markus and Pfeifer, Bastian and Kramer, Diether and Schulz, Stefan},
title = {Secondary Use of Clinical Problem List Entries for Neural Network-Based Disease Code Assignment},
journal = {Studies in Health Technology and Informatics},
year = {2023},
doi = {10.3233/shti230267}
}
EUVIMED