Description
This model maps extracted medical entities to SNOMED codes for the German language using sent_bert_base_cased (de) embeddings.
Predicted Entities
SNOMED Codes
Live Demo Open in Colab Copy S3 URI
How to use
documentAssembler = DocumentAssembler()\
.setInputCol("text")\
.setOutputCol("ner_chunk")
sbert_embedder = BertSentenceEmbeddings.pretrained("sent_bert_base_cased", "de")\
.setInputCols(["ner_chunk"])\
.setOutputCol("sbert_embeddings")
snomed_resolver = SentenceEntityResolverModel.pretrained("sbertresolve_snomed", "de", "clinical/models") \
.setInputCols(["sbert_embeddings"]) \
.setOutputCol("snomed_code")
snomed_pipelineModel = PipelineModel(
stages = [
documentAssembler,
sbert_embedder,
snomed_resolver])
snomed_lp = LightPipeline(snomed_pipelineModel)
val documentAssembler = DocumentAssembler()
.setInputCol("text")
.setOutputCol("ner_chunk")
val sbert_embedder = BertSentenceEmbeddings.pretrained("sent_bert_base_cased", "de")
.setInputCols("ner_chunk")
.setOutputCol("sbert_embeddings")
val snomed_resolver = SentenceEntityResolverModel.pretrained("sbertresolve_snomed", "de", "clinical/models") \
.setInputCols(["sbert_embeddings"]) \
.setOutputCol("snomed_code")
val snomed_pipelineModel = new PipelineModel().setStages(Array(documentAssembler, sbert_embedder, snomed_resolver))
val snomed_lp = LightPipeline(snomed_pipelineModel)
import nlu
nlu.load("de.resolve.snomed").predict("""Put your text here.""")
Results
| | chunks | code | resolutions | all_codes | all_distances |
|---:|:------------------|:--------|:-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------:|-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------:|:---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| 0 | Bronchialkarzinom | 22628 | Bronchialkarzinom, Bronchuskarzinom, Rektumkarzinom, Klavikulakarzinom, Lippenkarzinom, Urothelkarzinom, Hodenteratokarzinom, Unterbauchkarzinom, Teratokarzinom, Oropharynxkarzinom, Harnleiterkarzinom, Herzbeutelkarzinom, Thekazellkarzinom, Plattenepithelkarzinom, Weichteilkarzinom, Perikardkarzinom, Zervixkarzinom, Samenstrangkarzinom, Nierenkelchkarzinom, Querkolonkarzinom, Perianalkarzinom, Endozervixkarzinom, Parotiskarzinom, Gehörgangskarzinom, Prostatakarzinom| [22628, 111139, 18116, 107569, 18830, 22909, 16259, 111193, 22383, 19807, 22613, 20014, 74820, 21331, 30182, 20015, 23130, 22068, 20340, 29968, 15757, 23917, 25303, 17800, 21706] | [0.0000, 0.0073, 0.0090, 0.0098, 0.0098, 0.0102, 0.0102, 0.0110, 0.0111, 0.0120, 0.0121, 0.0123, 0.0128, 0.0130, 0.0129, 0.0131, 0.0128, 0.0131, 0.0135, 0.0133, 0.0137, 0.0137, 0.0139, 0.0137, 0.0139] |
Model Information
| Model Name: | sbertresolve_snomed |
| Compatibility: | Healthcare NLP 3.2.2+ |
| License: | Licensed |
| Edition: | Official |
| Input Labels: | [sentence_embeddings] |
| Output Labels: | [snomed_code] |
| Language: | de |
| Case sensitive: | false |