Description
This pipeline, extracts symptom entities in clinical text. It recognizes a comprehensive list of clinical symptoms.
How to use
from sparknlp.pretrained import PretrainedPipeline
pipeline = PretrainedPipeline("symptom_matcher_pipeline", "en", "clinical/models")
sample_text = """ The patient presents with severe headache and nausea for 3 days. She reports chest pain radiating to left arm with shortness of breath. Review of systems positive for fatigue, dizziness, and palpitations. Patient also complains of abdominal pain and constipation."""
result = pipeline.transform(spark.createDataFrame([[sample_text]]).toDF("text"))
from johnsnowlabs import nlp, medical
pipeline = nlp.PretrainedPipeline("symptom_matcher_pipeline", "en", "clinical/models")
sample_text = """ The patient presents with severe headache and nausea for 3 days. She reports chest pain radiating to left arm with shortness of breath. Review of systems positive for fatigue, dizziness, and palpitations. Patient also complains of abdominal pain and constipation."""
result = pipeline.transform(spark.createDataFrame([[sample_text]]).toDF("text"))
import com.johnsnowlabs.nlp.pretrained.PretrainedPipeline
val pipeline = PretrainedPipeline("symptom_matcher_pipeline", "en", "clinical/models")
val sample_text = """ The patient presents with severe headache and nausea for 3 days. She reports chest pain radiating to left arm with shortness of breath. Review of systems positive for fatigue, dizziness, and palpitations. Patient also complains of abdominal pain and constipation."""
val result = pipeline.transform(spark.createDataFrame([[sample_text]]).toDF("text"))
Results
| chunk | begin | end | label |
| :------------------ | ----: | --: | :------ |
| severe headache | 26 | 40 | SYMPTOM |
| nausea | 46 | 51 | SYMPTOM |
| chest pain | 77 | 86 | SYMPTOM |
| shortness of breath | 115 | 133 | SYMPTOM |
| fatigue | 167 | 173 | SYMPTOM |
| dizziness | 176 | 184 | SYMPTOM |
| palpitations | 191 | 202 | SYMPTOM |
| abdominal pain | 231 | 244 | SYMPTOM |
| constipation | 250 | 261 | SYMPTOM |
Model Information
| Model Name: | symptom_matcher_pipeline |
| Type: | pipeline |
| Compatibility: | Healthcare NLP 6.3.0+ |
| License: | Licensed |
| Edition: | Official |
| Language: | en |
| Size: | 963.9 KB |
Included Models
- DocumentAssembler
- SentenceDetector
- TokenizerModel
- TextMatcherInternalModel
- ChunkConverter