Pipeline for Detect Medication

Description

A pretrained pipeline to detect medication entities. It was built on the top of ner_posology_greedy model and also augmented with the drug names mentioned in UK and US drugbank datasets.

Predicted Entities

DRUG

Live Demo Open in Colab Download Copy S3 URI

How to use

from sparknlp.pretrained import PretrainedPipeline

ner_medication_pipeline = PretrainedPipeline("ner_medication_pipeline", "en", "clinical/models")

text = """The patient was prescribed metformin 1000 MG, and glipizide 2.5 MG. The other patient was given Fragmin 5000 units, Xenaderm to wounds topically b.i.d. and OxyContin 30 mg."""

result = ner_medication_pipeline.fullAnnotate([text])

import com.johnsnowlabs.nlp.pretrained.PretrainedPipeline

val ner_medication_pipeline = new PretrainedPipeline("ner_medication_pipeline", "en", "clinical/models")

val result = ner_medication_pipeline.fullAnnotate("The patient was prescribed metformin 1000 MG, and glipizide 2.5 MG. The other patient was given Fragmin 5000 units, Xenaderm to wounds topically b.i.d. and OxyContin 30 mg."")

Results

| ner_chunk          | entity   |
|:-------------------|:---------|
| metformin 1000 MG  | DRUG     |
| glipizide 2.5 MG   | DRUG     |
| Fragmin 5000 units | DRUG     |
| Xenaderm           | DRUG     |
| OxyContin 30 mg    | DRUG     |

Model Information

Included Models

• DocumentAssembler
• SentenceDetectorDLModel
• TokenizerModel
• WordEmbeddingsModel
• MedicalNerModel
• NerConverterInternalModel
• TextMatcherModel
• ChunkMergeModel
• Finisher