Clinical Assertion Negation Bert

🚀 Clinical Assertion / Negation Classification BERT

The Clinical Assertion and Negation Classification BERT classifies medical conditions in clinical patient letters into PRESENT, ABSENT, and POSSIBLE, helping to structure clinical information.

🚀 Quick Start

The Clinical Assertion and Negation Classification BERT is introduced in the paper Assertion Detection in Clinical Notes: Medical Language Models to the Rescue?. It is based on the ClinicalBERT - Bio + Discharge Summary BERT Model by Alsentzer et al. and fine - tuned on assertion data from the 2010 i2b2 challenge. The model helps structure information in clinical patient letters by classifying medical conditions mentioned in the letter into PRESENT, ABSENT and POSSIBLE.

💻 Usage Examples

Basic Usage

from transformers import AutoTokenizer, AutoModelForSequenceClassification, TextClassificationPipeline
tokenizer = AutoTokenizer.from_pretrained("bvanaken/clinical-assertion-negation-bert")
model = AutoModelForSequenceClassification.from_pretrained("bvanaken/clinical-assertion-negation-bert")

The model expects input in the form of spans/sentences with one marked entity to classify as PRESENT(0), ABSENT(1) or POSSIBLE(2). The entity in question is identified with the special token [entity] surrounding it.

Advanced Usage

input = "The patient recovered during the night and now denies any [entity] shortness of breath [entity]."

classifier = TextClassificationPipeline(model=model, tokenizer=tokenizer)

classification = classifier(input)
# [{'label': 'ABSENT', 'score': 0.9842607378959656}]

📚 Documentation

Cite

When working with the model, please cite our paper as follows:

@inproceedings{van-aken-2021-assertion,
    title = "Assertion Detection in Clinical Notes: Medical Language Models to the Rescue?",
    author = "van Aken, Betty  and
      Trajanovska, Ivana  and
      Siu, Amy  and
      Mayrdorfer, Manuel  and
      Budde, Klemens  and
      Loeser, Alexander",
    booktitle = "Proceedings of the Second Workshop on Natural Language Processing for Medical Conversations",
    year = "2021",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2021.nlpmc-1.5",
    doi = "10.18653/v1/2021.nlpmc-1.5"
}