Clinical Findings to UMLS Code Pipeline

Description

This pretrained pipeline maps entities (Clinical Findings) with their corresponding UMLS CUI codes. You’ll just feed your text and it will return the corresponding UMLS codes.

Live Demo Open in Colab Download Copy S3 URI

Available as Private API Endpoint

Deploy

Deploy

Deploy

How to use

from sparknlp.pretrained import PretrainedPipeline

pipeline = PretrainedPipeline("umls_clinical_findings_resolver_pipeline", "en", "clinical/models")
result = pipeline.fullAnnotate("HTG-induced pancreatitis associated with an acute hepatitis, and obesity.")

import com.johnsnowlabs.nlp.pretrained.PretrainedPipeline

val pipeline = PretrainedPipeline("umls_clinical_findings_resolver_pipeline", "en", "clinical/models")
val result = pipeline.annotate("HTG-induced pancreatitis associated with an acute hepatitis, and obesity")

import nlu
nlu.load("en.map_entity.umls_clinical_findings_resolver").predict("""HTG-induced pancreatitis associated with an acute hepatitis, and obesity""")

Results

+------------------------+---------+---------+
|chunk                   |ner_label|umls_code|
+------------------------+---------+---------+
|HTG-induced pancreatitis|PROBLEM  |C1963198 |
|an acute hepatitis      |PROBLEM  |C4750596 |
|obesity                 |PROBLEM  |C1963185 |
+------------------------+---------+---------+

Model Information

Included Models

• DocumentAssembler
• SentenceDetector
• TokenizerModel
• WordEmbeddingsModel
• MedicalNerModel
• NerConverter
• ChunkMapperModel
• ChunkMapperModel
• ChunkMapperFilterer
• Chunk2Doc
• BertSentenceEmbeddings
• SentenceEntityResolverModel
• ResolverMerger