Pipeline to Mapping UMLS Codes with Their Corresponding MESH Codes

Description

This pretrained pipeline is built on the top of umls_mesh_mapper model and maps UMLS codes to corresponding MESH codes

Copy S3 URI

How to use


from sparknlp.pretrained import PretrainedPipeline

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

sample_text = """ [['C0000530'], ['C0000726'], ['C0000343'], ['C5416820']]"""

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


from johnsnowlabs import nlp, medical

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

sample_text = """ [['C0000530'], ['C0000726'], ['C0000343'], ['C5416820']]"""

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


import com.johnsnowlabs.nlp.pretrained.PretrainedPipeline

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

val sample_text = """ [['C0000530'], ['C0000726'], ['C0000343'], ['C5416820']]"""

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

Results


| umls_code | mesh_code  |
| :-------- | :--------- |
| C0000530  | D015720    |
| C0000726  | D000005    |
| C0000343  | D015652    |
| C5416820  | C000722768 |

Model Information

Model Name: umls_mesh_mapping
Type: pipeline
Compatibility: Healthcare NLP 6.3.0+
License: Licensed
Edition: Official
Language: en
Size: 6.2 MB

Included Models

  • DocumentAssembler
  • Doc2Chunk
  • ChunkMapperModel