New Approaches in Mathematical and Data-Based Modeling for Newborn Screening

Newborn screening is a crucial health care program aiming to detect rare inherited metabolic diseases early. However, newborn screening for specific diseases faces challenges, such as false-positive screening results and a need for individualized disease management. In this research, new mathematical and data-based modeling approaches to support and improve newborn screening were developed. 

Therefore, data-based machine learning models using data from over two million newborns were developed to enhance diagnostic accuracy in newborn screening. Furthermore, a genome-based, mathematical whole-body model was developed to depict the metabolism of newborns and infants enabling personalized in silico simulations of newborns, and further developed to enable dynamic evaluations of infantile metabolism.