Description
Pretrained named entity recognition deep learning model for clinical terms. The SparkNLP deep learning model (NerDL) is inspired by a former state of the art model for NER: Chiu & Nicols, Named Entity Recognition with Bidirectional LSTM-CNN.
Predicted Entities
Problem, Test, Treatment
Live Demo Open in Colab Download
How to use
Use as part of an nlp pipeline with the following stages: DocumentAssembler, SentenceDetector, Tokenizer, WordEmbeddingsModel, NerDLModel. Add the NerConverter to the end of the pipeline to convert entity tokens into full entity chunks.
...
clinical_ner = NerDLModel.pretrained("ner_clinical_large", "en", "clinical/models") \
.setInputCols(["sentence", "token", "embeddings"]) \
.setOutputCol("ner")
...
nlpPipeline = Pipeline(stages=[document_assembler, sentence_detector, tokenizer, word_embeddings_clinical, clinical_ner, ner_converter])
model = nlpPipeline.fit(spark.createDataFrame([[""]]).toDF("text"))
results = model.transform(data)
...
val clinical_ner = NerDLModel.pretrained("ner_clinical_large", "en", "clinical/models")
.setInputCols(Array("sentence", "token", "embeddings"))
.setOutputCol("ner")
val nlpPipeline = new Pipeline().setStages(Array(document_assembler, sentence_detector, tokenizer, word_embeddings_clinical, clinical_ner, ner_converter))
val result = pipeline.fit(Seq.empty[String]).transform(data)
val results = LightPipeline(model).fullAnnotate(data)
Results
The output is a dataframe with a sentence per row and a `ner` column containing all of the entity labels in the sentence, entity character indices, and other metadata. To get only the tokens and entity labels, without the metadata, select `token.result` and `ner.result` from your output dataframe:
+-----------------------------------------------------------+---------+
|chunk |ner_label|
+-----------------------------------------------------------+---------+
|the G-protein-activated inwardly rectifying potassium (GIRK|TREATMENT|
|the genomicorganization |TREATMENT|
|a candidate gene forType II diabetes mellitus |PROBLEM |
|byapproximately |TREATMENT|
|single nucleotide polymorphisms |TREATMENT|
|aVal366Ala substitution |TREATMENT|
|an 8 base-pair |TREATMENT|
|insertion/deletion |PROBLEM |
|Ourexpression studies |TEST |
|the transcript in various humantissues |PROBLEM |
|fat andskeletal muscle |PROBLEM |
|furtherstudies |PROBLEM |
|the KCNJ9 protein |TREATMENT|
|evaluation |TEST |
|Type II diabetes |PROBLEM |
|the treatment |TREATMENT|
|breast cancer |PROBLEM |
|the standard therapy |TREATMENT|
|anthracyclines |TREATMENT|
|taxanes |TREATMENT|
+-----------------------------------------------------------+---------+
Model Information
| Model Name: | ner_clinical |
| Type: | ner |
| Compatibility: | Spark NLP 2.7.2+ |
| License: | Licensed |
| Edition: | Official |
| Input Labels: | [sentence, token, embeddings] |
| Output Labels: | [ner] |
| Language: | en |
| Dependencies: | embeddings_clinical |
Data Source
Trained on augmented 2010 i2b2 challenge data with ‘embeddings_clinical’. https://portal.dbmi.hms.harvard.edu/projects/n2c2-nlp/
Benchmarking
| | label | tp | fp | fn | prec | rec | f1 |
|---:|--------------:|------:|------:|------:|---------:|---------:|---------:|
| 0 | I-TREATMENT | 6625 | 1187 | 1329 | 0.848054 | 0.832914 | 0.840416 |
| 1 | I-PROBLEM | 15142 | 1976 | 2542 | 0.884566 | 0.856254 | 0.87018 |
| 2 | B-PROBLEM | 11005 | 1065 | 1587 | 0.911765 | 0.873968 | 0.892466 |
| 3 | I-TEST | 6748 | 923 | 1264 | 0.879677 | 0.842237 | 0.86055 |
| 4 | B-TEST | 8196 | 942 | 1029 | 0.896914 | 0.888455 | 0.892665 |
| 5 | B-TREATMENT | 8271 | 1265 | 1073 | 0.867345 | 0.885167 | 0.876165 |
| 6 | Macro-average | 55987 | 7358 | 8824 | 0.881387 | 0.863166 | 0.872181 |
| 7 | Micro-average | 55987 | 7358 | 8824 | 0.883842 | 0.86385 | 0.873732 |