Description
This pipeline is designed to extract all entities mappable to MESH codes.
How to use
from sparknlp.pretrained import PretrainedPipeline
ner_pipeline = PretrainedPipeline("ner_mesh_pipeline", "en", "clinical/models")
result = ner_pipeline.annotate("""
She was admitted to the hospital with chest pain and found to have bilateral pleural effusion, the right greater than the left.
We reviewed the pathology obtained from the pericardectomy in March 2006, which was diagnostic of mesothelioma.
At this time, chest tube placement for drainage of the fluid occurred and thoracoscopy with fluid biopsies,
which were performed, which revealed malignant mesothelioma.
""")
import com.johnsnowlabs.nlp.pretrained.PretrainedPipeline
val ner_pipeline = PretrainedPipeline("ner_mesh_pipeline", "en", "clinical/models")
val result = ner_pipeline.annotate("""
She was admitted to the hospital with chest pain and found to have bilateral pleural effusion, the right greater than the left.
We reviewed the pathology obtained from the pericardectomy in March 2006, which was diagnostic of mesothelioma.
At this time, chest tube placement for drainage of the fluid occurred and thoracoscopy with fluid biopsies,
which were performed, which revealed malignant mesothelioma.
""")
Results
| | chunks | begin | end | entities |
|---:|:---------------------------|--------:|------:|:-----------|
| 0 | chest pain | 39 | 48 | PROBLEM |
| 1 | bilateral pleural effusion | 68 | 93 | PROBLEM |
| 2 | the pathology | 142 | 154 | TEST |
| 3 | the pericardectomy | 170 | 187 | TREATMENT |
| 4 | mesothelioma | 228 | 239 | PROBLEM |
| 5 | chest tube placement | 257 | 276 | TREATMENT |
| 6 | drainage of the fluid | 282 | 302 | PROBLEM |
| 7 | thoracoscopy | 317 | 328 | TREATMENT |
| 8 | fluid biopsies | 335 | 348 | TEST |
| 9 | malignant mesothelioma | 389 | 410 | PROBLEM |
Model Information
| Model Name: | ner_mesh_pipeline |
| Type: | pipeline |
| Compatibility: | Healthcare NLP 6.0.2+ |
| License: | Licensed |
| Edition: | Official |
| Language: | en |
| Size: | 1.7 GB |
Included Models
- DocumentAssembler
- SentenceDetectorDLModel
- TokenizerModel
- WordEmbeddingsModel
- MedicalNerModel
- NerConverterInternalModel