Description
This is a pretrained named entity recognition deep learning model for clinical terminology. It is based on ner_jsl model, but with more generalised entities.
Predicted Entities
Death_Entity, Medical_Device, Vital_Sign, Alergen, Drug, Clinical_Dept, Lifestyle, Symptom, Body_Part, Physical_Measurement, Admission_Discharge, Date_Time, Age, Birth_Entity, Header, Oncological, Substance_Quantity, Test_Result, Test, Procedure, Treatment, Disease_Syndrome_Disorder, Pregnancy_Newborn, Demographics
Live Demo Open in Colab Download
How to use
embeddings_clinical = WordEmbeddingsModel().pretrained('embeddings_clinical', 'en', 'clinical/models') \
.setInputCols(['sentence', 'token']) \
.setOutputCol('embeddings')
clinical_ner = MedicalNerModel().pretrained("ner_jsl_slim", "en", "clinical/models") \
.setInputCols(["sentence", "token", "embeddings"]) \
.setOutputCol("ner")
...
nlpPipeline = Pipeline(stages=[document_assembler, sentence_detector, tokenizer, embeddings_clinical, clinical_ner, ner_converter])
model = nlpPipeline.fit(spark.createDataFrame([[""]]).toDF("text"))
results = model.transform(spark.createDataFrame([["HISTORY: 30-year-old female presents for digital bilateral mammography secondary to a soft tissue lump palpated by the patient in the upper right shoulder. The patient has a family history of breast cancer within her mother at age 58. Patient denies personal history of breast cancer."]], ["text"]))
...
val embeddings_clinical = WordEmbeddingsModel.pretrained("embeddings_clinical", "en", "clinical/models")
.setInputCols(["sentence", "token"])
.setOutputCol("embeddings")
val ner = MedicalNerModel.pretrained("ner_jsl_slim", "en", "clinical/models")
.setInputCols("sentence", "token", "embeddings")
.setOutputCol("ner")
...
val pipeline = new Pipeline().setStages(Array(document_assembler, sentence_detector, tokenizer, embeddings_clinical, ner, ner_converter))
val data = Seq("HISTORY: 30-year-old female presents for digital bilateral mammography secondary to a soft tissue lump palpated by the patient in the upper right shoulder. The patient has a family history of breast cancer within her mother at age 58. Patient denies personal history of breast cancer.").toDF("text")
val result = pipeline.fit(data).transform(data)
Results
| | chunk | entity |
|---:|:-----------------|:-------------|
| 0 | HISTORY: | Header |
| 1 | 30-year-old | Age |
| 2 | female | Demographics |
| 3 | mammography | Test |
| 4 | soft tissue lump | Symptom |
| 5 | shoulder | Body_Part |
| 6 | breast cancer | Oncological |
| 7 | her mother | Demographics |
| 8 | age 58 | Age |
| 9 | breast cancer | Oncological |
Model Information
| Model Name: | ner_jsl_slim |
| Compatibility: | Spark NLP for Healthcare 3.2.0+ |
| License: | Licensed |
| Edition: | Official |
| Input Labels: | [sentence, token, embeddings] |
| Output Labels: | [ner] |
| Language: | en |
Data Source
Trained on data annotated by JSL.
Benchmarking
|label |tp |fp |fn |prec |rec |f1 |
|---------------------------|----|----|----|----------|----------|----------|
|B-Medical_Device |2696|444 |282 |0.8585987 |0.90530556|0.88133377|
|I-Physical_Measurement |220 |16 |34 |0.9322034 |0.86614174|0.8979592 |
|B-Procedure |1800|239 |281 |0.8827857 |0.8649688 |0.8737864 |
|B-Drug |1865|218 |237 |0.89534324|0.88725024|0.8912784 |
|I-Test_Result |289 |203 |292 |0.58739835|0.49741825|0.5386766 |
|I-Pregnancy_Newborn |150 |41 |104 |0.7853403 |0.5905512 |0.6741573 |
|B-Admission_Discharge |255 |35 |6 |0.87931037|0.9770115 |0.92558986|
|B-Demographics |4609|119 |123 |0.9748308 |0.9740068 |0.97441864|
|I-Lifestyle |71 |49 |20 |0.59166664|0.7802198 |0.67298573|
|B-Header |2463|53 |122 |0.9789348 |0.9528046 |0.965693 |
|I-Date_Time |928 |184 |191 |0.8345324 |0.8293119 |0.83191395|
|B-Test_Result |866 |198 |262 |0.81390977|0.7677305 |0.79014593|
|I-Treatment |114 |37 |46 |0.7549669 |0.7125 |0.733119 |
|B-Clinical_Dept |688 |83 |76 |0.8923476 |0.90052354|0.8964169 |
|B-Test |1920|333 |313 |0.85219705|0.85982984|0.8559965 |
|B-Death_Entity |36 |9 |2 |0.8 |0.94736844|0.8674699 |
|B-Lifestyle |268 |58 |50 |0.8220859 |0.8427673 |0.8322981 |
|B-Date_Time |823 |154 |176 |0.8423746 |0.8238238 |0.83299595|
|I-Age |136 |34 |49 |0.8 |0.73513514|0.7661972 |
|I-Oncological |345 |41 |19 |0.8937824 |0.9478022 |0.91999996|
|I-Body_Part |3717|720 |424 |0.8377282 |0.8976093 |0.8666356 |
|B-Pregnancy_Newborn |153 |51 |104 |0.75 |0.5953307 |0.6637744 |
|B-Treatment |169 |41 |58 |0.8047619 |0.74449337|0.7734553 |
|I-Procedure |2302|326 |417 |0.8759513 |0.8466348 |0.8610435 |
|B-Birth_Entity |6 |5 |7 |0.54545456|0.46153846|0.5 |
|I-Vital_Sign |639 |197 |93 |0.76435405|0.8729508 |0.815051 |
|I-Header |4451|111 |216 |0.97566855|0.9537176 |0.9645682 |
|I-Death_Entity |2 |0 |0 |1 |1 |1 |
|I-Clinical_Dept |621 |54 |39 |0.92 |0.9409091 |0.9303371 |
|I-Test |1593|378 |353 |0.8082192 |0.81860226|0.81337756|
|B-Age |472 |43 |51 |0.91650486|0.90248567|0.9094413 |
|I-Symptom |4227|1271|1303|0.76882505|0.7643761 |0.7665941 |
|I-Demographics |321 |53 |53 |0.85828876|0.85828876|0.85828876|
|B-Body_Part |6312|912 |809 |0.87375414|0.88639235|0.8800279 |
|B-Physical_Measurement |91 |10 |17 |0.9009901 |0.8425926 |0.8708134 |
|B-Disease_Syndrome_Disorder|2817|336 |433 |0.8934348 |0.86676925|0.8799001 |
|B-Symptom |4522|830 |747 |0.8449178 |0.8582274 |0.8515206 |
|I-Disease_Syndrome_Disorder|2814|386 |530 |0.879375 |0.8415072 |0.86002445|
|I-Drug |3737|612 |517 |0.859278 |0.8784673 |0.8687667 |
|I-Medical_Device |1825|331 |131 |0.84647495|0.9330266 |0.8876459 |
|B-Oncological |276 |28 |27 |0.90789473|0.9108911 |0.9093904 |
|B-Vital_Sign |429 |97 |79 |0.81558937|0.8444882 |0.8297872 |
tp: 62038 fp: 9340 fn: 9110 labels: 46
Macro-average prec: 0.76782775, rec: 0.7648211, f1: 0.76632154
Micro-average prec: 0.86914736, rec: 0.87195706, f1: 0.87055