CIM10

The eds.cim10 pipeline component matches the CIM10 (French-language ICD) terminology.

Very low recall

When using the exact' matching mode, this component has a very poor recall performance. We can use thesimstring` mode to retrieve approximate matches, albeit at the cost of a significantly higher computation time.

Usage

import spacy

nlp = spacy.blank("fr")
nlp.add_pipe("eds.cim10", config=dict(term_matcher="simstring"))

text = "Le patient est suivi pour fièvres typhoïde et paratyphoïde."

doc = nlp(text)

doc.ents
# Out: (fièvres typhoïde et paratyphoïde,)

ent = doc.ents[0]

ent.label_
# Out: cim10

ent.kb_id_
# Out: A01

Configuration

The pipeline can be configured using the following parameters :

Parameter	Description	Default
term_matcher	Which algorithm should we use : exact or simstring	"exact"
term_matcher_config	Config of the algorithm (SimstringMatcher's for simstring)	{}
attr	spaCy attribute to match on (eg NORM, TEXT, LOWER)	"LOWER"
ignore_excluded	Whether to ignore excluded tokens for matching	False

Authors and citation

The eds.cim10 pipeline was developed by AP-HP's Data Science team.