de_ggponc_medbertde

Clinical NER model using spaCy's SpanCategorizer implementation and medBERT.de.

Usage:

!huggingface-cli download phlobo/de_ggponc_medbertde de_ggponc_medbertde-1.0.0-py3-none-any.whl --local-dir .
!pip install de_ggponc_medbertde-1.0.0-py3-none-any.whl

import spacy
nlp = spacy.load('de_ggponc_medbertde')
d = nlp("allein nach Versagen einer Behandlung mit Oxaliplatin und Irinotecan")
for e in d.spans['entities']:
  print(e, e.label_)

yields:

Oxaliplatin Clinical_Drug
Irinotecan Clinical_Drug
Versagen einer Behandlung Other_Finding
Behandlung mit Oxaliplatin und Irinotecan Therapeutic

The model has been trained on gold standard labels in GGPONC 2.0 (https://aclanthology.org/2022.lrec-1.389/).

It detects the following 8 entity classes:

• Findings: Diagnosis / Pathology and Other Findings
• Substances: Clinical Drug, Nutrients / Body Substances, External Substances
• Procedures: Therapeutic, Diagnostic

The configuration for training the model is available here: https://github.com/hpi-dhc/ggponc

When using the model, please cite the following publication:

@inproceedings{borchert-etal-2022-ggponc,
    title = "{GGPONC} 2.0 - The {G}erman Clinical Guideline Corpus for Oncology: Curation Workflow, Annotation Policy, Baseline {NER} Taggers",
    author = "Borchert, Florian  and
      Lohr, Christina  and
      Modersohn, Luise  and
      Witt, Jonas  and
      Langer, Thomas  and
      Follmann, Markus  and
      Gietzelt, Matthias  and
      Arnrich, Bert  and
      Hahn, Udo  and
      Schapranow, Matthieu-P.",
    booktitle = "Proceedings of the Thirteenth Language Resources and Evaluation Conference",
    month = jun,
    year = "2022",
    address = "Marseille, France",
    publisher = "European Language Resources Association",
    pages = "3650--3660"
}