Clinical Decision Support Tools

In collaboration with the UKBB Paediatric Pharmacology and Pharmacometrics Research Center, we have developed an online tool for the early detection of jaundice in neonates.

In a nutshell, we found that machine learning methods can accurately predict 48 hours in advance if a newborn is likely to receive a phototherapy treatment. Further information can be found in the related publication.

In collaboration with the University Children's Hospital Regensburg (KUNO) and Department of Pediatric Surgery and Pediatric Orthopedics at the Hospital St. Hedwig of the Order of St. John, we developed an online tool for predicting the diagnosis, management and severity of pediatric patients with suspected appendicitis. We utilize demographic, clinical, laboratory, scoring and ultrasound information to accurately predict all three target variables. Further details can be found in our Frontiers in Pediatrics paper.