Medizinische Universität Graz Austria/Österreich - Forschungsportal - Medical University of Graz
Gewählte Publikation:
SHR Neuro Krebs Kardio Lipid Stoffw Microb
Kreuzthaler, M; Pfeifer, B; Kramer, D; Schulz, S.
Secondary Use of Clinical Problem List Entries for Neural Network-Based Disease Code Assignment.
Stud Health Technol Inform. 2023; 302: 788-792.
Doi: 10.3233/SHTI230267
PubMed
FullText
FullText_MUG
- Führende Autor*innen der Med Uni Graz
- Kreuzthaler Markus Eduard
- Co-Autor*innen der Med Uni Graz
- Pfeifer Bastian
- Schulz Stefan
- Altmetrics:
- Dimensions Citations:
- Plum Analytics:
- Scite (citation analytics):
- Abstract:
- 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.
- Find related publications in this database (using NLM MeSH Indexing)
- Neural Networks, Computer - administration & dosage
- Language - administration & dosage
- International Classification of Diseases - administration & dosage
- Electronic Health Records - administration & dosage