Description
Named Entity Recognition annotators allow for a generic model to be trained by using a Deep Learning architecture (Char CNNs - BiLSTM - CRF - word embeddings) inspired on a former state of the art model for NER: Chiu & Nicols, Named Entity Recognition with Bidirectional LSTM, CNN.
Deidentification NER (Italian) is a Named Entity Recognition model that annotates text to find protected health information that may need to be de-identified. It detects 19 entities. This NER model is trained with a custom dataset internally annotated, a COVID-19 Italian de-identification research dataset making up 15% of the total data (Catelli et al.) and several data augmentation mechanisms.
Predicted Entities
DATE, AGE, SEX, PROFESSION, ORGANIZATION, PHONE, EMAIL, ZIP, STREET, CITY, COUNTRY, PATIENT, DOCTOR, HOSPITAL, MEDICALRECORD, SSN, IDNUM, USERNAME, URL
Live Demo Open in Colab Copy S3 URI
How to use
documentAssembler = DocumentAssembler()\
.setInputCol("text")\
.setOutputCol("document")
sentenceDetector = SentenceDetectorDLModel.pretrained("sentence_detector_dl", "xx")\
.setInputCols(["document"])\
.setOutputCol("sentence")
tokenizer = Tokenizer()\
.setInputCols(["sentence"])\
.setOutputCol("token")
embeddings = WordEmbeddingsModel.pretrained("w2v_cc_300d", "it")\
.setInputCols(["sentence", "token"])\
.setOutputCol("word_embeddings")
clinical_ner = MedicalNerModel.pretrained("ner_deid_subentity", "it", "clinical/models")\
.setInputCols(["sentence","token", "word_embeddings"])\
.setOutputCol("ner")
nlpPipeline = Pipeline(stages=[
documentAssembler,
sentenceDetector,
tokenizer,
embeddings,
clinical_ner])
text = ["Ho visto Gastone Montanariello (49 anni) riferito all' Ospedale San Camillo per diabete mal controllato con sintomi risalenti a marzo 2015."]
data = spark.createDataFrame([text]).toDF("text")
results = nlpPipeline.fit(data).transform(data)
val documentAssembler = new DocumentAssembler()
.setInputCol("text")
.setOutputCol("document")
val sentenceDetector = SentenceDetectorDLModel.pretrained("sentence_detector_dl", "xx")
.setInputCols(Array("document"))
.setOutputCol("sentence")
val tokenizer = new Tokenizer()
.setInputCols(Array("sentence"))
.setOutputCol("token")
val embeddings = WordEmbeddingsModel.pretrained("w2v_cc_300d", "it")
.setInputCols(Array("sentence", "token"))
.setOutputCol("embeddings")
val clinical_ner = MedicalNerModel.pretrained("ner_deid_subentity", "it", "clinical/models")
.setInputCols(Array("sentence","token","embeddings"))
.setOutputCol("ner")
val pipeline = new Pipeline().setStages(Array(
documentAssembler,
sentenceDetector,
tokenizer,
embeddings,
clinical_ner))
val text = "Ho visto Gastone Montanariello (49 anni) riferito all' Ospedale San Camillo per diabete mal controllato con sintomi risalenti a marzo 2015."
val data = Seq(text).toDF("text")
val results = pipeline.fit(data).transform(data)
import nlu
nlu.load("it.med_ner.deid_subentity").predict("""Ho visto Gastone Montanariello (49 anni) riferito all' Ospedale San Camillo per diabete mal controllato con sintomi risalenti a marzo 2015.""")
Results
+-------------+----------+
| token| ner_label|
+-------------+----------+
| Ho| O|
| visto| O|
| Gastone| B-PATIENT|
|Montanariello| I-PATIENT|
| (| O|
| 49| B-AGE|
| anni| O|
| )| O|
| riferito| O|
| all| O|
| '| O|
| Ospedale|B-HOSPITAL|
| San|I-HOSPITAL|
| Camillo|I-HOSPITAL|
| per| O|
| diabete| O|
| mal| O|
| controllato| O|
| con| O|
| sintomi| O|
| risalenti| O|
| a| O|
| marzo| B-DATE|
| 2015| I-DATE|
| .| O|
+-------------+----------+
Model Information
| Model Name: | ner_deid_subentity |
| Compatibility: | Healthcare NLP 3.4.2+ |
| License: | Licensed |
| Edition: | Official |
| Input Labels: | [sentence, token, embeddings] |
| Output Labels: | [ner] |
| Language: | it |
| Size: | 15.0 MB |
References
Benchmarking
label tp fp fn total precision recall f1
PATIENT 263.0 29.0 25.0 288.0 0.9007 0.9132 0.9069
HOSPITAL 365.0 36.0 48.0 413.0 0.9102 0.8838 0.8968
DATE 1164.0 13.0 26.0 1190.0 0.989 0.9782 0.9835
ORGANIZATION 72.0 25.0 26.0 98.0 0.7423 0.7347 0.7385
URL 41.0 0.0 0.0 41.0 1.0 1.0 1.0
CITY 421.0 9.0 19.0 440.0 0.9791 0.9568 0.9678
STREET 198.0 4.0 6.0 204.0 0.9802 0.9706 0.9754
USERNAME 20.0 2.0 2.0 22.0 0.9091 0.9091 0.9091
SEX 753.0 26.0 21.0 774.0 0.9666 0.9729 0.9697
IDNUM 113.0 3.0 7.0 120.0 0.9741 0.9417 0.9576
EMAIL 148.0 0.0 0.0 148.0 1.0 1.0 1.0
ZIP 148.0 3.0 1.0 149.0 0.9801 0.9933 0.9867
MEDICALRECORD 19.0 3.0 6.0 25.0 0.8636 0.76 0.8085
SSN 13.0 1.0 1.0 14.0 0.9286 0.9286 0.9286
PROFESSION 316.0 28.0 53.0 369.0 0.9186 0.8564 0.8864
PHONE 53.0 0.0 2.0 55.0 1.0 0.9636 0.9815
COUNTRY 182.0 14.0 15.0 197.0 0.9286 0.9239 0.9262
DOCTOR 769.0 77.0 62.0 831.0 0.909 0.9254 0.9171
AGE 763.0 8.0 18.0 781.0 0.9896 0.977 0.9832
macro - - - - - - 0.9328
micro - - - - - - 0.9494