Description
Pretrained named entity recognition deep learning model for clinical terms in German. The SparkNLP deep learning model (MedicalNerModel) is inspired by a former state of the art model for NER: Chiu & Nicols, Named Entity Recognition with Bidirectional LSTM-CNN.
Predicted Entities
PROBLEM, TEST, TREATMENT
Open in Colab Download Copy S3 URI
How to use
document_assembler = DocumentAssembler()\
.setInputCol("text")\
.setOutputCol("document")
sentence_detector = SentenceDetectorDLModel.pretrained("sentence_detector_dl", "xx") \
.setInputCols(["document"]) \
.setOutputCol("sentence")
tokenizer = Tokenizer()\
.setInputCols(["sentence"])\
.setOutputCol("token")
word_embeddings = WordEmbeddingsModel.pretrained("w2v_cc_300d", "de", "clinical/models")\
.setInputCols(["sentence", "token"])\
.setOutputCol("embeddings")
clinical_ner = MedicalNerModel.pretrained("ner_clinical", "de", "clinical/models") \
.setInputCols(["sentence", "token", "embeddings"]) \
.setOutputCol("ner")
ner_converter = NerConverterInternal()\
.setInputCols(["sentence", "token", "ner"])\
.setOutputCol("ner_chunk")
nlpPipeline = Pipeline(stages=[document_assembler,
sentence_detector,
tokenizer,
word_embeddings,
clinical_ner,
ner_converter])
model = nlpPipeline.fit(spark.createDataFrame([[""]]).toDF("text"))
sample_text= """Verschlechterung von Schmerzen oder Schwäche in den Beinen , Verlust der Darm - oder Blasenfunktion oder andere besorgniserregende Symptome.
Der Patient erhielt empirisch Ampicillin , Gentamycin und Flagyl sowie Narcan zur Umkehrung von Fentanyl .
ALT war 181 , AST war 156 , LDH war 336 , alkalische Phosphatase war 214 und Bilirubin war insgesamt 12,7 ."""
results = model.transform(spark.createDataFrame([[sample_text]], ["text"]))
val document_assembler = new DocumentAssembler()
.setInputCol("text")
.setOutputCol("document")
val sentence_detector = SentenceDetectorDLModel.pretrained("sentence_detector_dl", "xx")
.setInputCols("document")
.setOutputCol("sentence")
val tokenizer = new Tokenizer()
.setInputCols("sentence")
.setOutputCol("token")
val word_embeddings = WordEmbeddingsModel.pretrained("w2v_cc_300d", "de", "clinical/models")
.setInputCols(Array("sentence", "token"))
.setOutputCol("embeddings")
val ner = MedicalNerModel.pretrained("ner_clinical", "de", "clinical/models")
.setInputCols(Array("sentence", "token", "embeddings"))
.setOutputCol("ner")
val ner_converter = new NerConverterInternal()
.setInputCols(Array("sentence", "token", "ner"))
.setOutputCol("ner_chunk")
val pipeline = new Pipeline().setStages(Array(document_assembler, sentence_detector, tokenizer, word_embeddings, ner, ner_converter))
val data = Seq("""Verschlechterung von Schmerzen oder Schwäche in den Beinen , Verlust der Darm - oder Blasenfunktion oder andere besorgniserregende Symptome.
Der Patient erhielt empirisch Ampicillin , Gentamycin und Flagyl sowie Narcan zur Umkehrung von Fentanyl .
ALT war 181 , AST war 156 , LDH war 336 , alkalische Phosphatase war 214 und Bilirubin war insgesamt 12,7 .""").toDS().toDF("text")
val result = pipeline.fit(data).transform(data)
Results
+----------------------+---------+
|chunk |ner_label|
+----------------------+---------+
|Schmerzen |PROBLEM |
|Schwäche in den Beinen|PROBLEM |
|Verlust der Darm |PROBLEM |
|Blasenfunktion |PROBLEM |
|Symptome |PROBLEM |
|empirisch Ampicillin |TREATMENT|
|Gentamycin |TREATMENT|
|Flagyl |TREATMENT|
|Narcan |TREATMENT|
|Fentanyl |TREATMENT|
|ALT |TEST |
|AST |TEST |
|LDH |TEST |
|alkalische Phosphatase|TEST |
|Bilirubin |TEST |
+----------------------+---------+
Model Information
| Model Name: | ner_clinical |
| Compatibility: | Healthcare NLP 4.4.0+ |
| License: | Licensed |
| Edition: | Official |
| Input Labels: | [sentence, token, embeddings] |
| Output Labels: | [ner] |
| Language: | de |
| Size: | 2.0 MB |
Benchmarking
label precision recall f1-score support
B-PROBLEM 0.85 0.71 0.78 512
B-TEST 0.89 0.85 0.87 203
B-TREATMENT 0.84 0.82 0.83 238
I-PROBLEM 0.78 0.70 0.74 355
I-TEST 0.90 0.83 0.87 66
I-TREATMENT 0.62 0.71 0.66 75
O 0.94 0.97 0.95 4141
accuracy - - 0.91 5590
macro avg 0.83 0.80 0.81 5590
weighted avg 0.91 0.91 0.91 5590