Sentence Entity Resolver for ICD10-CM (sbiobert_base_cased_mli embeddings)

Description

This model maps extracted medical entities to ICD10-CM codes using sbiobert_base_cased_mli Sentence Bert Embeddings, and has faster load time, with a speedup of about 6X when compared to previous versions. Also the load process now is more memory friendly meaning that the maximum memory required during load time is smaller, reducing the chances of OOM exceptions, and thus relaxing hardware requirements.

Predicted Entities

Predicts ICD10-CM Codes and their normalized definitions.

Live Demo Open in Colab Copy S3 URI

How to use

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

sentence_detector = SentenceDetectorDLModel.pretrained("sentence_detector_dl_healthcare", "en", "clinical/models") \
	.setInputCols(["document"]) \
	.setOutputCol("sentence")

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

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

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

ner_converter = NerConverter() \
    .setInputCols(["sentence", "token", "ner"]) \
    .setOutputCol("entities")

chunk2doc = Chunk2Doc()\
    .setInputCols("entities")\
    .setOutputCol("ner_chunk_doc")

sbert_embedder = BertSentenceEmbeddings.pretrained("sbiobert_base_cased_mli","en","clinical/models")\
    .setInputCols(["ner_chunk_doc"])\
    .setOutputCol("sbert_embeddings")

icd10_resolver = SentenceEntityResolverModel.pretrained("sbiobertresolve_icd10cm","en", "clinical/models") \
    .setInputCols(["sbert_embeddings"]) \
    .setOutputCol("resolution")\
    .setDistanceFunction("EUCLIDEAN")

nlpPipeline = Pipeline(stages=[document_assembler, sentence_detector, tokenizer, word_embeddings, clinical_ner, ner_converter, chunk2doc, sbert_embedder, icd10_resolver])

data = spark.createDataFrame([["This is an 82 - year-old male with a history of prior tobacco use , hypertension , chronic renal insufficiency , COPD , gastritis , and TIA who initially presented to Braintree with a non-ST elevation MI and Guaiac positive stools , transferred to St . Margaret\'s Center for Women & Infants for cardiac catheterization with PTCA to mid LAD lesion complicated by hypotension and bradycardia requiring Atropine , IV fluids and transient dopamine possibly secondary to vagal reaction , subsequently transferred to CCU for close monitoring , hemodynamically stable at the time of admission to the CCU ."]]).toDF("text")

results = nlpPipeline.fit(data).transform(data)
val document_assembler = new DocumentAssembler()
	.setInputCol("text")
	.setOutputCol("document")

val sentence_detector = SentenceDetectorDLModel.pretrained("sentence_detector_dl_healthcare", "en", "clinical/models")
	.setInputCols("document")
	.setOutputCol("sentence")

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

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

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

val ner_converter = new NerConverter()
    .setInputCols(Array("sentence", "token", "ner"))
    .setOutputCol("entities")

val chunk2doc = new Chunk2Doc()
    .setInputCols("entities")
    .setOutputCol("ner_chunk_doc")

val sbert_embedder = BertSentenceEmbeddings.pretrained("sbiobert_base_cased_mli","en","clinical/models")
    .setInputCols("ner_chunk_doc")
    .setOutputCol("sbert_embeddings")

val icd10_resolver = SentenceEntityResolverModel.pretrained("sbiobertresolve_icd10cm","en", "clinical/models")
    .setInputCols(Array("sbert_embeddings"))
    .setOutputCol("resolution")
    .setDistanceFunction("EUCLIDEAN")

val pipeline = new Pipeline().setStages(Array(document_assembler, sentence_detector, tokenizer, word_embeddings, clinical_ner, ner_converter, chunk2doc, sbert_embedder, icd10_resolver))

val data = Seq("""This is an 82 - year-old male with a history of prior tobacco use , hypertension , chronic renal insufficiency , COPD , gastritis , and TIA who initially presented to Braintree with a non-ST elevation MI and Guaiac positive stools , transferred to St . Margaret\'s Center for Women & Infants for cardiac catheterization with PTCA to mid LAD lesion complicated by hypotension and bradycardia requiring Atropine , IV fluids and transient dopamine possibly secondary to vagal reaction , subsequently transferred to CCU for close monitoring , hemodynamically stable at the time of admission to the CCU .""").toDS().toDF("text")

val result = pipeline.fit(data).transform(data)
import nlu
nlu.load("en.resolve.icd10cm").predict("""This is an 82 - year-old male with a history of prior tobacco use , hypertension , chronic renal insufficiency , COPD , gastritis , and TIA who initially presented to Braintree with a non-ST elevation MI and Guaiac positive stools , transferred to St . Margaret\'s Center for Women & Infants for cardiac catheterization with PTCA to mid LAD lesion complicated by hypotension and bradycardia requiring Atropine , IV fluids and transient dopamine possibly secondary to vagal reaction , subsequently transferred to CCU for close monitoring , hemodynamically stable at the time of admission to the CCU .""")

Results

+--------------------+-----+---+---------+------+----------+--------------------+--------------------+
|               chunk|begin|end|   entity|  code|confidence|   all_k_resolutions|         all_k_codes|
+--------------------+-----+---+---------+------+----------+--------------------+--------------------+
|        hypertension|   68| 79|  PROBLEM|  I150|    0.2606|Renovascular hype...|I150:::K766:::I10...|
|chronic renal ins...|   83|109|  PROBLEM|  N186|    0.2059|End stage renal d...|N186:::D631:::P96...|
|                COPD|  113|116|  PROBLEM| I2781|    0.2132|Cor pulmonale (ch...|I2781:::J449:::J4...|
|           gastritis|  120|128|  PROBLEM| K5281|    0.1425|Eosinophilic gast...|K5281:::K140:::K9...|
|                 TIA|  136|138|  PROBLEM|  G459|    0.1152|Transient cerebra...|G459:::I639:::T79...|
|a non-ST elevatio...|  182|202|  PROBLEM|  I214|    0.0889|Non-ST elevation ...|I214:::I256:::M62...|
|Guaiac positive s...|  208|229|  PROBLEM|  K626|    0.0631|Ulcer of anus and...|K626:::K380:::R15...|
|cardiac catheteri...|  295|317|     TEST|  Z950|    0.2549|Presence of cardi...|Z950:::Z95811:::I...|
|                PTCA|  324|327|TREATMENT| Z9861|    0.1268|Coronary angiopla...|Z9861:::Z9862:::I...|
|      mid LAD lesion|  332|345|  PROBLEM|L02424|    0.1117|Furuncle of left ...|L02424:::Q202:::L...|
+--------------------+-----+---+---------+------+----------+--------------------+--------------------+

Model Information

Model Name: sbiobertresolve_icd10cm
Compatibility: Healthcare NLP 3.0.4+
License: Licensed
Edition: Official
Input Labels: [ner_chunk, sbert_embeddings]
Output Labels: [icd10cm_code]
Language: en
Case sensitive: false

Data Source

Trained on ICD10 Clinical Modification dataset with sbiobert_base_cased_mli sentence embeddings. https://www.icd10data.com/ICD10CM/Codes/