Detect Adverse Drug Events (bert-clinical)

Description

Detect adverse drug events in tweets, reviews, and medical text using pretrained NER model.

Predicted Entities

DRUG, ADE

Live Demo Open in Colab Copy S3 URI

How to use

document_assembler = DocumentAssembler()\
    .setInputCol("text")\
    .setOutputCol("document")

sentence_detector = SentenceDetector()\
    .setInputCols(["document"])\
    .setOutputCol("sentence")

tokenizer = Tokenizer()\
    .setInputCols(["sentence"])\
    .setOutputCol("token")

embeddings_clinical = BertEmbeddings.pretrained("biobert_clinical_base_cased")\
    .setInputCols(["sentence", "token"])\
    .setOutputCol("embeddings")

clinical_ner = MedicalNerModel.pretrained("ner_ade_clinicalbert", "en", "clinical/models")\
    .setInputCols(["sentence", "token", "embeddings"])\
    .setOutputCol("ner")

ner_converter = NerConverter() \
    .setInputCols(["sentence", "token", "ner"]) \
    .setOutputCol("ner_chunk")

nlpPipeline = Pipeline(stages=[document_assembler, sentence_detector, tokenizer, embeddings_clinical, clinical_ner, ner_converter])

model = nlpPipeline.fit(spark.createDataFrame([[""]]).toDF("text"))

results = model.transform(spark.createDataFrame([["EXAMPLE_TEXT"]]).toDF("text"))
val document_assembler = new DocumentAssembler()
    .setInputCol("text")
    .setOutputCol("document")
         
val sentence_detector = new SentenceDetector()
    .setInputCols("document")
    .setOutputCol("sentence")

val tokenizer = new Tokenizer()
    .setInputCols("sentence")
    .setOutputCol("token")

val embeddings_clinical = BertEmbeddings.pretrained("biobert_clinical_base_cased")
    .setInputCols(Array("sentence", "token"))
    .setOutputCol("embeddings")

val ner = MedicalNerModel.pretrained("ner_ade_clinicalbert", "en", "clinical/models")
    .setInputCols(Array("sentence", "token", "embeddings"))
    .setOutputCol("ner")

val ner_converter = new NerConverter()
 	.setInputCols(Array("sentence", "token", "ner"))
 	.setOutputCol("ner_chunk")

val pipeline = new Pipeline().setStages(Array(document_assembler, sentence_detector, tokenizer, embeddings_clinical, ner, ner_converter))

val result = pipeline.fit(data).transform(data)
import nlu
nlu.load("en.med_ner.ade.clinical_bert").predict("""Put your text here.""")

Model Information

Model Name: ner_ade_clinicalbert
Compatibility: Healthcare NLP 3.0.0+
License: Licensed
Edition: Official
Input Labels: [sentence, token, embeddings]
Output Labels: [ner]
Language: en

Benchmarking


label            precision    recall  f1-score   support
B-ADE               0.46      0.79      0.58      3582
B-DRUG              0.90      0.62      0.74     11763
I-ADE               0.45      0.76      0.56      4309
I-DRUG              0.96      0.26      0.41      7654
O                   0.96      0.98      0.97    303457
accuracy             -         -        0.94    330765
macro-avg           0.75      0.68      0.65    330765
weighted-avg        0.95      0.94      0.94    330765