Biomedical-clinical language model for Spanish

Table of contents

Click to expand

• Model description
• Intended uses and limitations
• How to use
• Limitations and bias
• Training
• Evaluation
• Additional information
  • Author
  • Contact information
  • Copyright
  • Licensing information
  • Funding
  • Citing information
  • Disclaimer

Model description

Biomedical pretrained language model for Spanish. For more details about the corpus, the pretraining and the evaluation, check the official repository.

Intended uses and limitations

The model is ready-to-use only for masked language modelling to perform the Fill Mask task (try the inference API or read the next section). However, it is intended to be fine-tuned on downstream tasks such as Named Entity Recognition or Text Classification.

How to use

Limitations and bias

At the time of submission, no measures have been taken to estimate the bias embedded in the model. However, we are well aware that our models may be biased since the corpora have been collected using crawling techniques on multiple web sources. We intend to conduct research in these areas in the future, and if completed, this model card will be updated.

Training

Tokenization and model pretraining

This model is a RoBERTa-based model trained on a biomedical-clinical corpus in Spanish collected from several sources (see next section). The training corpus has been tokenized using a byte version of Byte-Pair Encoding (BPE) used in the original RoBERTA model with a vocabulary size of 52,000 tokens. The pretraining consists of a masked language model training at the subword level following the approach employed for the RoBERTa base model with the same hyperparameters as in the original work. The training lasted a total of 48 hours with 16 NVIDIA V100 GPUs of 16GB DDRAM, using Adam optimizer with a peak learning rate of 0.0005 and an effective batch size of 2,048 sentences.

Training corpora and preprocessing

The training corpus is composed of several biomedical corpora in Spanish, collected from publicly available corpora and crawlers, and a real-world clinical corpus collected from more than 278K clinical documents and notes. To obtain a high-quality training corpus while retaining the idiosyncrasies of the clinical language, a cleaning pipeline has been applied only to the biomedical corpora, keeping the clinical corpus uncleaned. Essentially, the cleaning operations used are:

• data parsing in different formats
  • sentence splitting
  • language detection
  • filtering of ill-formed sentences
  • deduplication of repetitive contents
  • keep the original document boundaries

Then, the biomedical corpora are concatenated and further global deduplication among the biomedical corpora has been applied. Eventually, the clinical corpus is concatenated to the cleaned biomedical corpus resulting in a medium-size biomedical-clinical corpus for Spanish composed of more than 1B tokens. The table below shows some basic statistics of the individual cleaned corpora:

Evaluation

The model has been fine-tuned on three Named Entity Recognition (NER) tasks using three clinical NER datasets:

• PharmaCoNER: is a track on chemical and drug mention recognition from Spanish medical texts (for more info see: https://temu.bsc.es/pharmaconer/).

• CANTEMIST: is a shared task specifically focusing on named entity recognition of tumor morphology, in Spanish (for more info see: https://zenodo.org/record/3978041#.YTt5qH2xXbQ).

• ICTUSnet: consists of 1,006 hospital discharge reports of patients admitted for stroke from 18 different Spanish hospitals. It contains more than 79,000 annotations for 51 different kinds of variables.

We addressed the NER task as a token classification problem using a standard linear layer along with the BIO tagging schema. We compared our models with the general-domain Spanish roberta-base-bne, the general-domain multilingual model that supports Spanish mBERT, the domain-specific English model BioBERT, and three domain-specific models based on continual pre-training, mBERT-Galén, XLM-R-Galén and BETO-Galén. The table below shows the F1 scores obtained:

The fine-tuning scripts can be found in the official GitHub repository.

Additional information

Author

Text Mining Unit (TeMU) at the Barcelona Supercomputing Center (bsc-temu@bsc.es)

Contact information

For further information, send an email to plantl-gob-es@bsc.es

Copyright

Copyright by the Spanish State Secretariat for Digitalization and Artificial Intelligence (SEDIA) (2022)

Licensing information

Apache License, Version 2.0

Funding

This work was funded by the Spanish State Secretariat for Digitalization and Artificial Intelligence (SEDIA) within the framework of the Plan-TL.

Citing information

If you use these models, please cite our work:

@inproceedings{carrino-etal-2022-pretrained,
    title = "Pretrained Biomedical Language Models for Clinical {NLP} in {S}panish",
    author = "Carrino, Casimiro Pio  and
      Llop, Joan  and
      P{\`a}mies, Marc  and
      Guti{\'e}rrez-Fandi{\~n}o, Asier  and
      Armengol-Estap{\'e}, Jordi  and
      Silveira-Ocampo, Joaqu{\'\i}n  and
      Valencia, Alfonso  and
      Gonzalez-Agirre, Aitor  and
      Villegas, Marta",
    booktitle = "Proceedings of the 21st Workshop on Biomedical Language Processing",
    month = may,
    year = "2022",
    address = "Dublin, Ireland",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2022.bionlp-1.19",
    doi = "10.18653/v1/2022.bionlp-1.19",
    pages = "193--199",
    abstract = "This work presents the first large-scale biomedical Spanish language models trained from scratch, using large biomedical corpora consisting of a total of 1.1B tokens and an EHR corpus of 95M tokens. We compared them against general-domain and other domain-specific models for Spanish on three clinical NER tasks. As main results, our models are superior across the NER tasks, rendering them more convenient for clinical NLP applications. Furthermore, our findings indicate that when enough data is available, pre-training from scratch is better than continual pre-training when tested on clinical tasks, raising an exciting research question about which approach is optimal. Our models and fine-tuning scripts are publicly available at HuggingFace and GitHub.",
}

Disclaimer

Click to expand

The models published in this repository are intended for a generalist purpose and are available to third parties. These models may have bias and/or any other undesirable distortions.

When third parties, deploy or provide systems and/or services to other parties using any of these models (or using systems based on these models) or become users of the models, they should note that it is their responsibility to mitigate the risks arising from their use and, in any event, to comply with applicable regulations, including regulations regarding the use of Artificial Intelligence.

In no event shall the owner of the models (SEDIA – State Secretariat for Digitalization and Artificial Intelligence) nor the creator (BSC – Barcelona Supercomputing Center) be liable for any results arising from the use made by third parties of these models.

Los modelos publicados en este repositorio tienen una finalidad generalista y están a disposición de terceros. Estos modelos pueden tener sesgos y/u otro tipo de distorsiones indeseables.

Cuando terceros desplieguen o proporcionen sistemas y/o servicios a otras partes usando alguno de estos modelos (o utilizando sistemas basados en estos modelos) o se conviertan en usuarios de los modelos, deben tener en cuenta que es su responsabilidad mitigar los riesgos derivados de su uso y, en todo caso, cumplir con la normativa aplicable, incluyendo la normativa en materia de uso de inteligencia artificial.

En ningún caso el propietario de los modelos (SEDIA – Secretaría de Estado de Digitalización e Inteligencia Artificial) ni el creador (BSC – Barcelona Supercomputing Center) serán responsables de los resultados derivados del uso que hagan terceros de estos modelos.