Edit on GitHub
  • John Snow Labs
  • Jun 12, 2026

Mapping Entities (Drug Substance) with Corresponding UMLS CUI Codes

  • en
  • chunk_mapper
  • licensed
  • clinical
  • umls
  • drug_substance

Description

This model maps drug substance entities extracted by NER to UMLS CUI codes covering Drug Substance semantic types (T121, T131, T195, T200), comprising approximately 736,000 name–CUI pairs. It is trained on the 2026AA release of the Unified Medical Language System (UMLS) dataset.

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

ner_model_converter = NerConverterInternal()\
    .setInputCols(["sentence", "token", "clinical_ner"])\
    .setOutputCol("ner_chunk")

chunkerMapper = ChunkMapperModel.pretrained("umls_drug_substance_mapper", "en", "clinical/models")\
    .setInputCols(["ner_chunk"])\
    .setOutputCol("mappings")\
    .setRels(["umls_code"])\
    .setLowerCase(True)

mapper_pipeline = Pipeline(stages=[
    document_assembler,
    sentence_detector,
    tokenizer,
    word_embeddings,
    ner_model,
    ner_model_converter,
    chunkerMapper,
])
data = spark.createDataFrame([["The patient was given metformin, lenvatinib and sumatriptan 50 MG Oral Tablet. She was prescribed penicillin and aspirin for treatment."]]).toDF("text")
result = mapper_pipeline.fit(data).transform(data)



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

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

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

word_embeddings = nlp.WordEmbeddingsModel.pretrained("embeddings_clinical", "en", "clinical/models")\
    .setInputCols(["sentence", "token"])\
    .setOutputCol("embeddings")

ner_model = medical.NerModel.pretrained("ner_posology_greedy", "en", "clinical/models")\
    .setInputCols(["sentence", "token", "embeddings"])\
    .setOutputCol("clinical_ner")

ner_model_converter = medical.NerConverterInternal()\
    .setInputCols(["sentence", "token", "clinical_ner"])\
    .setOutputCol("ner_chunk")

chunkerMapper = medical.ChunkMapperModel.pretrained("umls_drug_substance_mapper", "en", "clinical/models")\
    .setInputCols(["ner_chunk"])\
    .setOutputCol("mappings")\
    .setRels(["umls_code"])\
    .setLowerCase(True)

mapper_pipeline = nlp.Pipeline(stages=[
    document_assembler,
    sentence_detector,
    tokenizer,
    word_embeddings,
    ner_model,
    ner_model_converter,
    chunkerMapper,
])
data = spark.createDataFrame([["The patient was given metformin, lenvatinib and sumatriptan 50 MG Oral Tablet. She was prescribed penicillin and aspirin for treatment."]]).toDF("text")
result = mapper_pipeline.fit(data).transform(data)



val documentAssembler = new DocumentAssembler()
    .setInputCol("text")
    .setOutputCol("document")

val sentenceDetector = new SentenceDetector()
    .setInputCols(Array("document"))
    .setOutputCol("sentence")

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

val wordEmbeddings = WordEmbeddingsModel.pretrained("embeddings_clinical", "en", "clinical/models")
    .setInputCols(Array("sentence", "token"))
    .setOutputCol("embeddings")

val nerModel = MedicalNerModel.pretrained("ner_posology_greedy", "en", "clinical/models")
    .setInputCols(Array("sentence", "token", "embeddings"))
    .setOutputCol("clinical_ner")

val nerModelConverter = new NerConverterInternal()
    .setInputCols(Array("sentence", "token", "clinical_ner"))
    .setOutputCol("ner_chunk")

val chunkerMapper = ChunkMapperModel.pretrained("umls_drug_substance_mapper", "en", "clinical/models")
    .setInputCols(Array("ner_chunk"))
    .setOutputCol("mappings")
    .setRels(Array("umls_code"))
    .setLowerCase(true)

val mapperPipeline = new Pipeline().setStages(Array(
    documentAssembler,
    sentenceDetector,
    tokenizer,
    wordEmbeddings,
    nerModel,
    nerModelConverter,
    chunkerMapper,
))

val data = Seq("The patient was given metformin, lenvatinib and sumatriptan 50 MG Oral Tablet. She was prescribed penicillin and aspirin for treatment.").toDF("text")
val result = mapperPipeline.fit(data).transform(data)

Results

| ner_chunk                     | umls_code   |
|:------------------------------|:------------|
| metformin                     | C0025598    |
| lenvatinib                    | C2986924    |
| sumatriptan 50 MG Oral Tablet | C0980430    |
| penicillin                    | C0030842    |
| aspirin                       | C0004057    |

Model Information

Model Name: umls_drug_substance_mapper
Compatibility: Healthcare NLP 6.4.0+
License: Licensed
Edition: Official
Input Labels: [ner_chunk]
Output Labels: [mappings]
Language: en
Size: 30.3 MB