In Acta paediatrica (Oslo, Norway : 1992)
AIM : Sepsis is a leading cause of morbidity and mortality in neonates. Early diagnosis is key but difficult due to non-specific signs. We investigate the predictive value of machine learning-assisted analysis of non-invasive, high frequency monitoring data and demographic factors to detect neonatal sepsis.
METHODS : Single center study, including a representative cohort of 325 infants (2866 hospitalization days). Personalized event timelines including interventions and clinical findings were generated. Time-domain features from heart rate, respiratory rate and oxygen saturation values were calculated and demographic factors included. Sepsis prediction was performed using Naïve-Bayes algorithm in a maximum a posteriori framework up to 24 hours before clinical sepsis suspicion.
RESULTS : Twenty sepsis cases were identified. Combining multiple vital signs improved algorithm performance compared to heart rate characteristics alone. This enabled a prediction of sepsis with an area under the receiver operating characteristics curve of 0.82, up to 24 hours before clinical sepsis suspicion. Moreover, 10 hours prior to clinical suspicion, the risk of sepsis increased 150-fold.
CONCLUSION : The present algorithm using non-invasive patient data provides useful predictive value for neonatal sepsis detection. Machine learning-assisted algorithms are promising novel methods that could help individualize patient care and reduce morbidity and mortality.
Honoré Antoine, Forsberg David, Adolphson Katja, Chatterjee Saikat, Jost Kerstin, Herlenius Eric
2023-Jan-06
Artificial intelligence, Clinical Decision Support system, Naïve-Bayes classifier, physiological monitoring, prediction, respiration-related
