Pipeline to Extract Anatomical Entities from Oncology Texts

Description

This pretrained pipeline is built on the top of ner_oncology_anatomy_general model.

Predicted Entities

Anatomical_Site, Direction

Copy S3 URI

How to use

from sparknlp.pretrained import PretrainedPipeline

pipeline = PretrainedPipeline("ner_oncology_anatomy_general_pipeline", "en", "clinical/models")

text = '''The patient presented a mass in her left breast, and a possible metastasis in her lungs and in her liver.'''

result = pipeline.fullAnnotate(text)
import com.johnsnowlabs.nlp.pretrained.PretrainedPipeline

val pipeline = new PretrainedPipeline("ner_oncology_anatomy_general_pipeline", "en", "clinical/models")

val text = "The patient presented a mass in her left breast, and a possible metastasis in her lungs and in her liver."

val result = pipeline.fullAnnotate(text)

Results

|    | ner_chunks   |   begin |   end | ner_label       |   confidence |
|---:|:-------------|--------:|------:|:----------------|-------------:|
|  0 | left         |      36 |    39 | Direction       |       0.9825 |
|  1 | breast       |      41 |    46 | Anatomical_Site |       0.9005 |
|  2 | lungs        |      82 |    86 | Anatomical_Site |       0.9735 |
|  3 | liver        |      99 |   103 | Anatomical_Site |       0.9817 |

Model Information

Model Name: ner_oncology_anatomy_general_pipeline
Type: pipeline
Compatibility: Healthcare NLP 4.3.0+
License: Licensed
Edition: Official
Language: en
Size: 1.7 GB

Included Models

  • DocumentAssembler
  • SentenceDetectorDLModel
  • TokenizerModel
  • WordEmbeddingsModel
  • MedicalNerModel
  • NerConverterInternalModel