Last week, we released Spark NLP 2.6 which was the 14th release of the open-source library in 2020. We also found that PyPI downloads of Spark NLP have doubled to 200,000/month in just two months – from June to August.
This week, we are releasing Spark NLP for Healthcare 2.6 – the new major version of the commercial library. In addition to all the goodies of the latest open-source package, new features include:
Clinical Relation Extraction
This release comes with a new, deep-learning-based, and trainable algorithm for relation extraction. Five pre-trained models are bundled:
This Colab notebook shows these models in action – from detecting temporal relationships between clinical events to drug-drug interactions:
Clinical Named Entity Recognition
Three new out-of-the-box NER models are now available:
ner_human_phenotype_gene_clinical (genes & phenotypes)
ner_human_phenotype_go_clinical (also normalized to the gene ontology)
ner_chemprot_clinical (chemical-protein interactions)
This Colab notebook shows these models in action.
Clinical & Legal Models in German
Three new pre-trained models are now available in the German language:
- German Clinical NER model for 19 clinical entities
- German Legal NER model for 19 legal entities
- German Clinical Entity Normalization to ICD-10GM
Out-of-the-box Clinical Pipelines
For the first time, this release includes three pretrained clinical pipelines. Pretrained pipelines are already fitted using certain annotators and transformers according to various use cases – for example, combining entity recognition, assertion status detection, and relation extraction as simply as:
pipeline = PretrainedPipeline('explain_clinical_doc_carp', 'en', 'clinical/models')
pipeline.annotate(‘my clinical record’)
Start a free trial now to try out these new capabilities – or expect an email with upgrade details soon if you’re already a customer.
If you found this Article interesting, why not review the other Articles in our archive.