Description
This pretrained model maps drugs with their corresponding Adverse Drug Events.
Predicted Entities
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")
word_embeddings = WordEmbeddingsModel.pretrained("embeddings_clinical", "en", "clinical/models")\
.setInputCols(["sentence", "token"])\
.setOutputCol("embeddings")
#NER model to detect drug in the text
ner = MedicalNerModel.pretrained('ner_posology_greedy', 'en', 'clinical/models') \
.setInputCols(["sentence", "token", "embeddings"]) \
.setOutputCol("ner")
ner_chunk = NerConverter() \
.setInputCols(["sentence", "token", "ner"]) \
.setOutputCol("ner_chunk")\
chunkMapper = ChunkMapperModel.pretrained("drug_ade_mapper", "en", "clinical/models")\
.setInputCols(["ner_chunk"])\
.setOutputCol("mappings")\
.setRels(["ADE"])
pipeline = Pipeline().setStages([document_assembler,
sentence_detector,
tokenizer,
word_embeddings,
ner,
ner_chunk,
chunkMapper])
text = ["""The patient was prescribed 1000 mg fish oil and multivitamins.
She was discharged on zopiclone and ambrisentan"""]
data = spark.createDataFrame([text]).toDF("text")
result= pipeline.fit(data).transform(data)
val document_assembler = new DocumentAssembler()
.setInputCol("text")
.setOutputCol("document")
val sentence_detector = new SentenceDetector()
.setInputCols(Array("document"))
.setOutputCol("sentence")
val tokenizer = new Tokenizer()
.setInputCols("sentence")
.setOutputCol("token")
val word_embeddings = WordEmbeddingsModel.pretrained("embeddings_clinical", "en", "clinical/models")
.setInputCols(Array("sentence", "token"))
.setOutputCol("embeddings")
#NER model to detect drug in the text
val ner = MedicalNerModel.pretrained("ner_posology_greedy", "en", "clinical/models")
.setInputCols(Array("sentence", "token", "embeddings"))
.setOutputCol("ner")
val ner_chunk = new NerConverter()
.setInputCols(Array("sentence", "token", "ner"))
.setOutputCol("ner_chunk")
val chunkMapper = ChunkMapperModel.pretrained("drug_ade_mapper", "en", "clinical/models")
.setInputCols(Array("ner_chunk"))
.setOutputCol("mappings")
.setRels(Array("ADE"))
val pipeline = new Pipeline(stages = Array(
document_assembler,
sentence_detector,
tokenizer,
word_embeddings,
ner,
ner_chunk,
chunkMapper))
val data = Seq("The patient was prescribed 1000 mg fish oil and multivitamins. She was discharged on zopiclone and ambrisentan").toDS.toDF("text")
val result= pipeline.fit(data).transform(data)
import nlu
nlu.load("en.map_entity.drug_ade").predict("""The patient was prescribed 1000 mg fish oil and multivitamins.
She was discharged on zopiclone and ambrisentan""")
Results
+----------------+------------+-------------------------------------------------------------------------------------------+
|ner_chunk |ade_mappings|all_relations |
+----------------+------------+-------------------------------------------------------------------------------------------+
|1000 mg fish oil|Dizziness |Myocardial infarction:::Nausea |
|multivitamins |Erythema |Acne:::Dry skin:::Skin burning sensation:::Inappropriate schedule of product administration|
|zopiclone |Vomiting |Malaise:::Drug interaction:::Asthenia:::Hyponatraemia |
|ambrisentan |Dyspnoea |Therapy interrupted:::Death:::Dizziness:::Drug ineffective |
+----------------+------------+-------------------------------------------------------------------------------------------+
Model Information
| Model Name: | drug_ade_mapper |
| Compatibility: | Healthcare NLP 4.0.2+ |
| License: | Licensed |
| Edition: | Official |
| Input Labels: | [ner_pos_chunk] |
| Output Labels: | [mappings] |
| Language: | en |
| Size: | 7.9 MB |
References
Data from the FDA Adverse Event Reporting System (FAERS) for the years 2020, 2021 and 2022 were used as the source for this mapper model.
https://fis.fda.gov/extensions/FPD-QDE-FAERS/FPD-QDE-FAERS.html