Description
This pretrained pipeline maps SNOMED codes to ICD10CM codes without using any text data. You’ll just feed a comma or white space-delimited SNOMED codes and it will return the corresponding candidate ICD10CM codes as a list (multiple ICD10 codes for each Snomed code). For the time being, it supports 132K Snomed codes and 30K ICD10 codes and will be augmented & enriched in the next releases.
How to use
from sparknlp.pretrained import PretrainedPipeline
pipeline = PretrainedPipeline( "snomed_icd10cm_mapping","en","clinical/models")
pipeline.annotate('721617001 733187009 109006')
import com.johnsnowlabs.nlp.pretrained.PretrainedPipeline
val pipeline = new PretrainedPipeline("icd10cm_snomed_mapping","en","clinical/models")
val result = pipeline.annotate('721617001 733187009 109006')
import nlu
nlu.load("en.map_entity.snomed_to_icd10cm.pipe").predict("""721617001 733187009 109006""")
Results
{'snomed': ['721617001', '733187009', '109006'],
'icd10cm': ['K22.70, C15.5',
'M89.59, M89.50, M96.89',
'F41.9, F40.10, F94.8, F93.0, F40.8, F93.8']}
Model Information
| Model Name: | snomed_icd10cm_mapping |
| Type: | pipeline |
| Compatibility: | Healthcare NLP 3.0.2+ |
| License: | Licensed |
| Edition: | Official |
| Language: | en |
Included Models
- DocumentAssembler
- TokenizerModel
- LemmatizerModel
- Finisher