Description
This specialized radiology pipeline can;
-
extract radiology related entities,
-
assign assertion status to the extracted entities,
-
establish relations between the extracted entities from the clinical documents.
In this pipeline, five NER, one assertion and one relation extraction model were used to achive those tasks.
-
Clinical Entity Labels:
Imaging_Test,Imaging_Technique,ImagingFindings,OtherFindings,BodyPart,Direction,Test,Symptom,Disease_Syndrome_Disorder,Medical_Device,Procedure,Measurements,Units,Gender,Metastasis,Invasion,Route,Treatment,Drug,Form,Frequency,Dosage,Date,Test_Result,Medical_Device -
Assertion Status Labels:
Confirmed,Suspected,Negative -
Relation Extraction Labels:
is_related,not_related
Predicted Entities
BodyPart, Date, Direction, Disease_Syndrome_Disorder, Dosage, Drug, Duration, Form, Frequency, Gender, ImagingFindings, Imaging_Technique, Imaging_Test, Invasion, ManualFix, Measurements, Medical_Device, Metastasis, OtherFindings, Procedure, RelativeDate, Route, Symptom, Test, Test_Result, Treatment, Units
How to use
from sparknlp.pretrained import PretrainedPipeline
ner_pipeline = PretrainedPipeline("explain_clinical_doc_radiology", "en", "clinical/models")
result = ner_pipeline.annotate("""Bilateral breast ultrasound was subsequently performed, which demonstrated an ovoid mass measuring approximately 0.5 x 0.5 x 0.4 cm in diameter located within the anteromedial aspect of the left shoulder. This mass demonstrates isoechoic echotexture to the adjacent muscle, with no evidence of internal color flow. This may represent benign fibrous tissue or a lipoma.""")
import com.johnsnowlabs.nlp.pretrained.PretrainedPipeline
val ner_pipeline = PretrainedPipeline("explain_clinical_doc_radiology", "en", "clinical/models")
val result = ner_pipeline.annotate("""Bilateral breast ultrasound was subsequently performed, which demonstrated an ovoid mass measuring approximately 0.5 x 0.5 x 0.4 cm in diameter located within the anteromedial aspect of the left shoulder. This mass demonstrates isoechoic echotexture to the adjacent muscle, with no evidence of internal color flow. This may represent benign fibrous tissue or a lipoma.""")
Results
# NER Result
| | sentence_id | chunks | begin | end | entities |
|---:|--------------:|:-----------------------------------------|--------:|------:|:--------------------------|
| 0 | 0 | Bilateral breast | 0 | 15 | BodyPart |
| 1 | 0 | ultrasound | 17 | 26 | Imaging_Test |
| 2 | 0 | ovoid mass | 78 | 87 | ImagingFindings |
| 3 | 0 | 0.5 x 0.5 x 0.4 cm | 113 | 130 | Measurements |
| 4 | 0 | anteromedial aspect of the left shoulder | 163 | 202 | BodyPart |
| 5 | 1 | mass | 210 | 213 | ImagingFindings |
| 6 | 1 | isoechoic echotexture | 228 | 248 | ImagingFindings |
| 7 | 1 | adjacent muscle | 257 | 271 | BodyPart |
| 8 | 1 | internal color flow | 294 | 312 | ImagingFindings |
| 9 | 2 | benign fibrous tissue | 334 | 354 | ImagingFindings |
| 10 | 2 | lipoma | 361 | 366 | Disease_Syndrome_Disorder |
# Assertion Result
| | sentence_id | chunks | begin | end | entities | assertion |
|---:|--------------:|:----------------------|--------:|------:|:--------------------------|:------------|
| 0 | 0 | ovoid mass | 78 | 87 | ImagingFindings | Confirmed |
| 1 | 1 | mass | 210 | 213 | ImagingFindings | Confirmed |
| 2 | 1 | isoechoic echotexture | 228 | 248 | ImagingFindings | Confirmed |
| 3 | 1 | internal color flow | 294 | 312 | ImagingFindings | Negative |
| 4 | 2 | benign fibrous tissue | 334 | 354 | ImagingFindings | Suspected |
| 5 | 2 | lipoma | 361 | 366 | Disease_Syndrome_Disorder | Suspected |
# Relation Extraction Result
| | sentence | entity1_begin | entity1_end | chunk1 | entity1 | entity2_begin | entity2_end | chunk2 | entity2 | relation | confidence |
|---:|-----------:|----------------:|--------------:|:----------------------|:----------------|----------------:|--------------:|:-------------------|:----------------|:-----------|-------------:|
| 0 | 0 | 0 | 15 | Bilateral breast | BodyPart | 17 | 26 | ultrasound | Imaging_Test | is_related | 1 |
| 1 | 0 | 0 | 15 | Bilateral breast | BodyPart | 78 | 87 | ovoid mass | ImagingFindings | is_related | 0.999997 |
| 2 | 0 | 17 | 26 | ultrasound | Imaging_Test | 78 | 87 | ovoid mass | ImagingFindings | is_related | 0.999569 |
| 3 | 0 | 78 | 87 | ovoid mass | ImagingFindings | 113 | 130 | 0.5 x 0.5 x 0.4 cm | Measurements | is_related | 1 |
| 4 | 1 | 210 | 213 | mass | ImagingFindings | 257 | 271 | adjacent muscle | BodyPart | is_related | 0.997639 |
| 5 | 1 | 228 | 248 | isoechoic echotexture | ImagingFindings | 257 | 271 | adjacent muscle | BodyPart | is_related | 0.999999 |
Model Information
| Model Name: | explain_clinical_doc_radiology |
| Type: | pipeline |
| Compatibility: | Healthcare NLP 5.2.1+ |
| License: | Licensed |
| Edition: | Official |
| Language: | en |
| Size: | 1.8 GB |
Included Models
- DocumentAssembler
- SentenceDetectorDLModel
- TokenizerModel
- WordEmbeddingsModel
- MedicalNerModel
- NerConverterInternalModel
- MedicalNerModel
- NerConverterInternalModel
- MedicalNerModel
- NerConverterInternalModel
- MedicalNerModel
- NerConverterInternalModel
- MedicalNerModel
- NerConverterInternalModel
- ChunkMergeModel
- ChunkMergeModel
- AssertionDLModel
- PerceptronModel
- DependencyParserModel
- RelationExtractionModel