Pipeline to Mapping SNOMED Codes with Their Corresponding UMLS Codes

Description

This pretrained pipeline is built on the top of snomed_umls_mapper model and maps SNOMED codes to corresponding UMLS codes.

Copy S3 URI

How to use


from sparknlp.pretrained import PretrainedPipeline

pipeline = PretrainedPipeline("snomed_umls_mapping", "en", "clinical/models")

sample_text = """ [['acebutolol'], ['fluids']]"""

result = pipeline.transform(spark.createDataFrame(sample_text).toDF("text"))


from johnsnowlabs import nlp, medical

pipeline = nlp.PretrainedPipeline("snomed_umls_mapping", "en", "clinical/models")

sample_text = """ [['acebutolol'], ['fluids']]"""

result = pipeline.transform(spark.createDataFrame(sample_text).toDF("text"))


import com.johnsnowlabs.nlp.pretrained.PretrainedPipeline

val pipeline = PretrainedPipeline("snomed_umls_mapping", "en", "clinical/models")

val sample_text = """ [['acebutolol'], ['fluids']]"""

val result = pipeline.transform(spark.createDataFrame(sample_text).toDF("text"))

Results


| chunk      | snomed_code | umls_code |
| :--------- | ----------: | :-------- |
| acebutolol |    68088000 | C0000946  |
| fluids     |   118431008 | C1289919  |

Model Information

Model Name: snomed_umls_mapping
Type: pipeline
Compatibility: Healthcare NLP 6.3.0+
License: Licensed
Edition: Official
Language: en
Size: 692.6 MB

Included Models

  • DocumentAssembler
  • BertSentenceEmbeddings
  • SentenceEntityResolverModel
  • Resolution2Chunk
  • ChunkMapperModel