AI and the World's Hard Problems · Medicine & drug discovery · Nature (computational biology) · 2026-10-26
Development and validation of an interpretable machine learning model for early prediction of refractory mycoplasma pneumoniae pneumonia: a multi-center study
Early prediction of refractory Mycoplasma pneumoniae pneumonia (RMPP) in children is difficult with conventional methods. Using data from 6,587 children across four centers, the authors compared six machine learning models, and a random forest performed best. In two external cohorts it achieved AUROCs of 0.774 and 0.807, with fever duration before admission, neutrophil percentage and age as the leading predictors.