Skip to content

edsnlp.pipelines.ner.scores.sofa.patterns

regex = ['\\bsofa\\b'] module-attribute

digits = '[^\\d]*(\\d*)' module-attribute

value_extract = [dict(name='method_max', regex='(max)', reduce_mode='keep_first'), dict(name='method_24h', regex='(24h)', reduce_mode='keep_first'), dict(name='method_adm', regex='(admission)', reduce_mode='keep_first'), dict(name='value', regex='^.*?[\\n\\W]*?(\\d+)(?![h0-9])')] module-attribute

score_normalization_str = 'score_normalization.sofa' module-attribute

score_normalization(extracted_score)

Sofa score normalization. If available, returns the integer value of the SOFA score.

Source code in edsnlp/pipelines/ner/scores/sofa/patterns.py 
34
35
36
37
38
39
40
41
42
43
@spacy.registry.misc(score_normalization_str)
def score_normalization(extracted_score: Union[str, None]):

    """
    Sofa score normalization.
    If available, returns the integer value of the SOFA score.
    """
    score_range = list(range(0, 30))
    if (extracted_score is not None) and (int(extracted_score) in score_range):
        return int(extracted_score)