Heart rate characteristics and demographic factors have long been used to aid early detection of late-onset sepsis, however respiratory data may contain additional signatures of infection.
In this episode we meet Early Career Investigator Brynne Sullivan from the University of Virginia. She and her team developed machine learning models to predict late-onset sepsis that were trained on heart rate and respiratory data to provide a cardiorespiratory early warning system which outperformed models using heart rate or demographics alone.
Read the full article here: Cardiorespiratory signature of neonatal sepsis: development and validation of prediction models in 3 NICUs | Pediatric Research
","author_name":"Nature Publishing Group"}