Description
This pipeline can be used to extract test mentions in medical text.
Predicted Entities
TEST
How to use
from sparknlp.pretrained import PretrainedPipeline
ner_pipeline = PretrainedPipeline("ner_test_benchmark_pipeline", "en", "clinical/models")
text = """PHYSICAL EXAMINATION :
On physical examination , the patient was a well developed , stocky gentleman .
The blood pressure was 115/80 , pulse 80 , respirations of 20 , venous pressure elevated at 3 cm above the clavicle at 90 degrees .
There were very small , barely palpable carotid pulses .
There was dullness at the right base , with a high diaphragm and possibly some fluid .
The cardiac examination showed a left ventricular tap at the fifth intercostal space left of the midclavicular line .
There was a grade II / VI systolic ejection murmur in the aortic area , no third sound , and paradoxical splitting of the second sound .
The liver was not palpable .
There were diminished pulses in the legs .
LABORATORY DATA :
The hemoglobin was 14.4 grams percent , white blood count 6,900 , platelet count 125,000 , sodium 137 mEq. per liter , potassium of 4.7 , BUN and creatinine of 23 and 1.3 mg percent .
The electrocardiogram showed left ventricular hypertrophy and non-specific ST-T wave changes .
The chest film showed massive cardiomegaly with pulmonary venous engorgement ."""
result = ner_pipeline.fullAnnotate(text)
from sparknlp.pretrained import PretrainedPipeline
ner_pipeline = nlp.PretrainedPipeline("ner_test_benchmark_pipeline", "en", "clinical/models")
text = """PHYSICAL EXAMINATION :
On physical examination , the patient was a well developed , stocky gentleman .
The blood pressure was 115/80 , pulse 80 , respirations of 20 , venous pressure elevated at 3 cm above the clavicle at 90 degrees .
There were very small , barely palpable carotid pulses .
There was dullness at the right base , with a high diaphragm and possibly some fluid .
The cardiac examination showed a left ventricular tap at the fifth intercostal space left of the midclavicular line .
There was a grade II / VI systolic ejection murmur in the aortic area , no third sound , and paradoxical splitting of the second sound .
The liver was not palpable .
There were diminished pulses in the legs .
LABORATORY DATA :
The hemoglobin was 14.4 grams percent , white blood count 6,900 , platelet count 125,000 , sodium 137 mEq. per liter , potassium of 4.7 , BUN and creatinine of 23 and 1.3 mg percent .
The electrocardiogram showed left ventricular hypertrophy and non-specific ST-T wave changes .
The chest film showed massive cardiomegaly with pulmonary venous engorgement ."""
result = ner_pipeline.fullAnnotate(text)
import com.johnsnowlabs.nlp.pretrained.PretrainedPipeline
val ner_pipeline = PretrainedPipeline("ner_test_benchmark_pipeline", "en", "clinical/models")
val text = """PHYSICAL EXAMINATION :
On physical examination , the patient was a well developed , stocky gentleman .
The blood pressure was 115/80 , pulse 80 , respirations of 20 , venous pressure elevated at 3 cm above the clavicle at 90 degrees .
There were very small , barely palpable carotid pulses .
There was dullness at the right base , with a high diaphragm and possibly some fluid .
The cardiac examination showed a left ventricular tap at the fifth intercostal space left of the midclavicular line .
There was a grade II / VI systolic ejection murmur in the aortic area , no third sound , and paradoxical splitting of the second sound .
The liver was not palpable .
There were diminished pulses in the legs .
LABORATORY DATA :
The hemoglobin was 14.4 grams percent , white blood count 6,900 , platelet count 125,000 , sodium 137 mEq. per liter , potassium of 4.7 , BUN and creatinine of 23 and 1.3 mg percent .
The electrocardiogram showed left ventricular hypertrophy and non-specific ST-T wave changes .
The chest film showed massive cardiomegaly with pulmonary venous engorgement ."""
val result = ner_pipeline.fullAnnotate(text)
Results
| | chunk | begin | end | ner_label |
|---:|:---------------------|--------:|------:|:------------|
| 0 | PHYSICAL EXAMINATION | 0 | 19 | TEST |
| 1 | physical examination | 26 | 45 | TEST |
| 2 | blood pressure | 107 | 120 | TEST |
| 3 | pulse | 135 | 139 | TEST |
| 4 | respirations | 146 | 157 | TEST |
| 5 | venous pressure | 167 | 181 | TEST |
| 6 | pulses | 283 | 288 | TEST |
| 7 | cardiac examination | 383 | 401 | TEST |
| 8 | pulses | 685 | 690 | TEST |
| 9 | hemoglobin | 728 | 737 | TEST |
| 10 | white blood count | 764 | 780 | TEST |
| 11 | platelet count | 790 | 803 | TEST |
| 12 | sodium | 815 | 820 | TEST |
| 13 | mEq | 826 | 828 | TEST |
| 14 | potassium | 843 | 851 | TEST |
| 15 | BUN | 862 | 864 | TEST |
| 16 | creatinine | 870 | 879 | TEST |
| 17 | electrocardiogram | 912 | 928 | TEST |
| 18 | chest film | 1007 | 1016 | TEST |
Model Information
| Model Name: | ner_test_benchmark_pipeline |
| Type: | pipeline |
| Compatibility: | Healthcare NLP 5.5.3+ |
| License: | Licensed |
| Edition: | Official |
| Language: | en |
| Size: | 1.8 GB |
Included Models
- DocumentAssembler
- SentenceDetector
- TokenizerModel
- WordEmbeddingsModel
- TextMatcherInternalModel
- TextMatcherInternalModel
- MedicalNerModel
- NerConverterInternalModel
- MedicalNerModel
- NerConverterInternalModel
- MedicalNerModel
- NerConverterInternalModel
- MedicalNerModel
- NerConverterInternalModel
- MedicalNerModel
- NerConverterInternalModel
- ChunkMergeModel
- ChunkMergeModel
Benchmarking
label precision recall f1-score support
O 0.995 0.996 0.996 79006
TEST 0.914 0.897 0.906 3565
accuracy - - 0.992 82571
macro-avg 0.955 0.946 0.951 82571
weighted-avg 0.992 0.992 0.992 82571