Description
Pretrained named entity recognition deep learning model for clinical terminology. The SparkNLP deep learning model (NerDL) is inspired by a former state of the art model for NER: Chiu & Nicols, Named Entity Recognition with Bidirectional LSTM-CNN.
Predicted Entities
Injury_or_Poisoning, Direction, Test, Admission_Discharge, Death_Entity, Relationship_Status, Duration, Respiration, Hyperlipidemia, Birth_Entity, Age, Labour_Delivery, Family_History_Header, BMI, Temperature, Alcohol, Kidney_Disease, Oncological, Medical_History_Header, Cerebrovascular_Disease, Oxygen_Therapy, O2_Saturation, Psychological_Condition, Heart_Disease, Employment, Obesity, Disease_Syndrome_Disorder, Pregnancy, ImagingFindings, Procedure, Medical_Device, Race_Ethnicity, Section_Header, Symptom, Treatment, Substance, Route, Drug_Ingredient, Blood_Pressure, Diet, I-Age, External_body_part_or_region, LDL, VS_Finding, Allergen, EKG_Findings, Imaging_Technique, I-Diet, Triglycerides, RelativeTime, Gender, Pulse, Social_History_Header, Substance_Quantity, Diabetes, Modifier, Internal_organ_or_component, Clinical_Dept, Form, Drug_BrandName, Strength, Fetus_NewBorn, RelativeDate, Height, Test_Result, Sexually_Active_or_Sexual_Orientation, Frequency, Time, Weight, Vaccine, Vital_Signs_Header, Communicable_Disease, Dosage, Overweight, Hypertension, HDL, Total_Cholesterol, Smoking, Date.
Live Demo Open in Colab Copy S3 URI
How to use
Use as part of an nlp pipeline with the following stages: DocumentAssembler, SentenceDetector, Tokenizer, WordEmbeddingsModel, NerDLModel. Add the NerConverter to the end of the pipeline to convert entity tokens into full entity chunks.
...
embeddings_clinical = WordEmbeddingsModel.pretrained('embeddings_clinical', 'en', 'clinical/models') \
.setInputCols(['sentence', 'token']) \
.setOutputCol('embeddings')
clinical_ner = NerDLModel.pretrained("jsl_ner_wip_clinical", "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([["The patient is a 21-day-old Caucasian male here for 2 days of congestion - mom has been suctioning yellow discharge from the patient's nares, plus she has noticed some mild problems with his breathing while feeding (but negative for any perioral cyanosis or retractions). One day ago, mom also noticed a tactile temperature and gave the patient Tylenol. Baby also has had some decreased p.o. intake. His normal breast-feeding is down from 20 minutes q.2h. to 5 to 10 minutes secondary to his respiratory congestion. He sleeps well, but has been more tired and has been fussy over the past 2 days. The parents noticed no improvement with albuterol treatments given in the ER. His urine output has also decreased; normally he has 8 to 10 wet and 5 dirty diapers per 24 hours, now he has down to 4 wet diapers per 24 hours. Mom denies any diarrhea. His bowel movements are yellow colored and soft in nature."]], ["text"]))
...
val embeddings_clinical = WordEmbeddingsModel.pretrained("embeddings_clinical", "en", "clinical/models")
.setInputCols("sentence", "token")
.setOutputCol("embeddings")
val ner = NerDLModel.pretrained("jsl_ner_wip_clinical", "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("The patient is a 21-day-old Caucasian male here for 2 days of congestion - mom has been suctioning yellow discharge from the patient's nares, plus she has noticed some mild problems with his breathing while feeding (but negative for any perioral cyanosis or retractions). One day ago, mom also noticed a tactile temperature and gave the patient Tylenol. Baby also has had some decreased p.o. intake. His normal breast-feeding is down from 20 minutes q.2h. to 5 to 10 minutes secondary to his respiratory congestion. He sleeps well, but has been more tired and has been fussy over the past 2 days. The parents noticed no improvement with albuterol treatments given in the ER. His urine output has also decreased; normally he has 8 to 10 wet and 5 dirty diapers per 24 hours, now he has down to 4 wet diapers per 24 hours. Mom denies any diarrhea. His bowel movements are yellow colored and soft in nature.").toDF("text")
val result = pipeline.fit(data).transform(data)
import nlu
nlu.load("en.med_ner").predict("""The patient is a 21-day-old Caucasian male here for 2 days of congestion - mom has been suctioning yellow discharge from the patient's nares, plus she has noticed some mild problems with his breathing while feeding (but negative for any perioral cyanosis or retractions). One day ago, mom also noticed a tactile temperature and gave the patient Tylenol. Baby also has had some decreased p.o. intake. His normal breast-feeding is down from 20 minutes q.2h. to 5 to 10 minutes secondary to his respiratory congestion. He sleeps well, but has been more tired and has been fussy over the past 2 days. The parents noticed no improvement with albuterol treatments given in the ER. His urine output has also decreased; normally he has 8 to 10 wet and 5 dirty diapers per 24 hours, now he has down to 4 wet diapers per 24 hours. Mom denies any diarrhea. His bowel movements are yellow colored and soft in nature.""")
Results
The output is a dataframe with a sentence per row and a "ner" column containing all of the entity labels in the sentence, entity character indices, and other metadata. To get only the tokens and entity labels, without the metadata, select "token.result" and "ner.result" from your output dataframe or add the "Finisher" to the end of your pipeline.
+-----------------------------------------+----------------------------+
|chunk |ner_label |
+-----------------------------------------+----------------------------+
|21-day-old |Age |
|Caucasian |Race_Ethnicity |
|male |Gender |
|for 2 days |Duration |
|congestion |Symptom |
|mom |Gender |
|yellow |Modifier |
|discharge |Symptom |
|nares |External_body_part_or_region|
|she |Gender |
|mild |Modifier |
|problems with his breathing while feeding|Symptom |
|perioral cyanosis |Symptom |
|retractions |Symptom |
|One day ago |RelativeDate |
|mom |Gender |
|Tylenol |Drug_BrandName |
|Baby |Age |
|decreased p.o. intake |Symptom |
|His |Gender |
+-----------------------------------------+----------------------------+
Model Information
| Model Name: | jsl_ner_wip_clinical |
| Type: | ner |
| Compatibility: | Spark NLP 2.7.0+ |
| Edition: | Official |
| License: | Licensed |
| Input Labels: | [sentence,token, embeddings] |
| Output Labels: | [ner] |
| Language: | [en] |
| Case sensitive: | false |
Data Source
Trained on data gathered and manually annotated by John Snow Labs. https://www.johnsnowlabs.com/data/
Benchmarking
entity tp fp fn total precision recall f1
VS_Finding 235.0 46.0 43.0 278.0 0.8363 0.8453 0.8408
Direction 3972.0 465.0 458.0 4430.0 0.8952 0.8966 0.8959
Respiration 82.0 4.0 4.0 86.0 0.9535 0.9535 0.9535
Cerebrovascular_D... 93.0 20.0 24.0 117.0 0.823 0.7949 0.8087
Family_History_He... 88.0 6.0 3.0 91.0 0.9362 0.967 0.9514
Heart_Disease 447.0 82.0 119.0 566.0 0.845 0.7898 0.8164
RelativeTime 158.0 80.0 59.0 217.0 0.6639 0.7281 0.6945
Strength 624.0 58.0 53.0 677.0 0.915 0.9217 0.9183
Smoking 121.0 11.0 4.0 125.0 0.9167 0.968 0.9416
Medical_Device 3716.0 491.0 466.0 4182.0 0.8833 0.8886 0.8859
Pulse 136.0 22.0 14.0 150.0 0.8608 0.9067 0.8831
Psychological_Con... 135.0 9.0 29.0 164.0 0.9375 0.8232 0.8766
Overweight 2.0 1.0 0.0 2.0 0.6667 1.0 0.8
Triglycerides 3.0 0.0 2.0 5.0 1.0 0.6 0.75
Obesity 42.0 5.0 6.0 48.0 0.8936 0.875 0.8842
Admission_Discharge 318.0 24.0 11.0 329.0 0.9298 0.9666 0.9478
HDL 3.0 0.0 0.0 3.0 1.0 1.0 1.0
Diabetes 110.0 14.0 8.0 118.0 0.8871 0.9322 0.9091
Section_Header 3740.0 148.0 157.0 3897.0 0.9619 0.9597 0.9608
Age 627.0 75.0 48.0 675.0 0.8932 0.9289 0.9107
O2_Saturation 34.0 14.0 17.0 51.0 0.7083 0.6667 0.6869
Kidney_Disease 96.0 12.0 34.0 130.0 0.8889 0.7385 0.8067
Test 2504.0 545.0 498.0 3002.0 0.8213 0.8341 0.8276
Communicable_Disease 21.0 10.0 6.0 27.0 0.6774 0.7778 0.7241
Hypertension 162.0 5.0 10.0 172.0 0.9701 0.9419 0.9558
External_body_par... 2626.0 356.0 413.0 3039.0 0.8806 0.8641 0.8723
Oxygen_Therapy 81.0 15.0 14.0 95.0 0.8438 0.8526 0.8482
Modifier 2341.0 404.0 539.0 2880.0 0.8528 0.8128 0.8324
Test_Result 1007.0 214.0 255.0 1262.0 0.8247 0.7979 0.8111
BMI 9.0 1.0 0.0 9.0 0.9 1.0 0.9474
Labour_Delivery 57.0 23.0 33.0 90.0 0.7125 0.6333 0.6706
Employment 271.0 59.0 55.0 326.0 0.8212 0.8313 0.8262
Fetus_NewBorn 66.0 33.0 51.0 117.0 0.6667 0.5641 0.6111
Clinical_Dept 923.0 110.0 83.0 1006.0 0.8935 0.9175 0.9053
Time 29.0 13.0 16.0 45.0 0.6905 0.6444 0.6667
Procedure 3185.0 462.0 501.0 3686.0 0.8733 0.8641 0.8687
Diet 36.0 20.0 45.0 81.0 0.6429 0.4444 0.5255
Oncological 459.0 61.0 55.0 514.0 0.8827 0.893 0.8878
LDL 3.0 0.0 3.0 6.0 1.0 0.5 0.6667
Symptom 7104.0 1302.0 1200.0 8304.0 0.8451 0.8555 0.8503
Temperature 116.0 6.0 8.0 124.0 0.9508 0.9355 0.9431
Vital_Signs_Header 215.0 29.0 24.0 239.0 0.8811 0.8996 0.8903
Relationship_Status 49.0 2.0 1.0 50.0 0.9608 0.98 0.9703
Total_Cholesterol 11.0 4.0 5.0 16.0 0.7333 0.6875 0.7097
Blood_Pressure 158.0 18.0 22.0 180.0 0.8977 0.8778 0.8876
Injury_or_Poisoning 579.0 130.0 127.0 706.0 0.8166 0.8201 0.8184
Drug_Ingredient 1716.0 153.0 132.0 1848.0 0.9181 0.9286 0.9233
Treatment 136.0 36.0 60.0 196.0 0.7907 0.6939 0.7391
Pregnancy 123.0 36.0 51.0 174.0 0.7736 0.7069 0.7387
Vaccine 13.0 2.0 6.0 19.0 0.8667 0.6842 0.7647
Disease_Syndrome_... 2981.0 559.0 446.0 3427.0 0.8421 0.8699 0.8557
Height 30.0 10.0 15.0 45.0 0.75 0.6667 0.7059
Frequency 595.0 99.0 138.0 733.0 0.8573 0.8117 0.8339
Route 858.0 76.0 89.0 947.0 0.9186 0.906 0.9123
Duration 351.0 99.0 108.0 459.0 0.78 0.7647 0.7723
Death_Entity 43.0 14.0 5.0 48.0 0.7544 0.8958 0.819
Internal_organ_or... 6477.0 972.0 991.0 7468.0 0.8695 0.8673 0.8684
Alcohol 80.0 18.0 13.0 93.0 0.8163 0.8602 0.8377
Substance_Quantity 6.0 7.0 4.0 10.0 0.4615 0.6 0.5217
Date 498.0 38.0 19.0 517.0 0.9291 0.9632 0.9459
Hyperlipidemia 47.0 3.0 3.0 50.0 0.94 0.94 0.94
Social_History_He... 99.0 7.0 7.0 106.0 0.934 0.934 0.934
Race_Ethnicity 116.0 0.0 0.0 116.0 1.0 1.0 1.0
Imaging_Technique 40.0 18.0 47.0 87.0 0.6897 0.4598 0.5517
Drug_BrandName 859.0 62.0 61.0 920.0 0.9327 0.9337 0.9332
RelativeDate 566.0 124.0 143.0 709.0 0.8203 0.7983 0.8091
Gender 6096.0 80.0 101.0 6197.0 0.987 0.9837 0.9854
Dosage 244.0 31.0 57.0 301.0 0.8873 0.8106 0.8472
Form 234.0 32.0 55.0 289.0 0.8797 0.8097 0.8432
Medical_History_H... 114.0 9.0 10.0 124.0 0.9268 0.9194 0.9231
Birth_Entity 4.0 2.0 3.0 7.0 0.6667 0.5714 0.6154
Substance 59.0 8.0 11.0 70.0 0.8806 0.8429 0.8613
Sexually_Active_o... 5.0 3.0 4.0 9.0 0.625 0.5556 0.5882
Weight 90.0 10.0 21.0 111.0 0.9 0.8108 0.8531
macro - - - - - - 0.8148
micro - - - - - - 0.8788