Reason

The eds.reason pipeline uses a rule-based algorithm to detect spans that relate to the reason of the hospitalisation. It was designed at AP-HP's EDS.

Usage

The following snippet matches a simple terminology, and looks for spans of hospitalisation reasons. It is complete and can be run as is.

import spacy

text = """COMPTE RENDU D'HOSPITALISATION du 11/07/2018 au 12/07/2018
MOTIF D'HOSPITALISATION
Monsieur Dupont Jean Michel, de sexe masculin, âgée de 39 ans, née le 23/11/1978, a été
hospitalisé du 11/08/2019 au 17/08/2019 pour attaque d'asthme.

ANTÉCÉDENTS
Antécédents médicaux :
Premier épisode d'asthme en mai 2018."""

nlp = spacy.blank("fr")

# Extraction of entities
nlp.add_pipe(
    "eds.matcher",
    config=dict(
        terms=dict(
            respiratoire=[
                "asthmatique",
                "asthme",
                "toux",
            ]
        )
    ),
)


nlp.add_pipe("eds.normalizer")
nlp.add_pipe("eds.reason", config=dict(use_sections=True))
doc = nlp(text)

reason = doc.spans["reasons"][0]
reason
# Out: hospitalisé du 11/08/2019 au 17/08/2019 pour attaque d'asthme.

reason._.is_reason
# Out: True

entities = reason._.ents_reason
entities
# Out: [asthme]

entities[0].label_
# Out: 'respiratoire'

ent = entities[0]
ent._.is_reason
# Out: True

Configuration

The pipeline can be configured using the following parameters :

Parameter	Explanation	Default
reasons	Reasons patterns	None (use pre-defined patterns)
attr	spaCy attribute to match on, eg NORM or TEXT	"NORM"
use_sections	Whether to use sections	False
ignore_excluded	Whether to ignore excluded tokens	False

Declared extensions

The eds.reason pipeline adds the key reasons to doc.spans and declares one spaCy extension, on the Span objects called ents_reason.

The ents_reason extension is a list of named entities that overlap the Span, typically entities found in previous pipelines like matcher.

It also declares the boolean extension is_reason. This extension is set to True for the Reason Spans but also for the entities that overlap the reason span.

Authors and citation

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