Preview

PDF, English - main document Download (8MB) | Terms of use

Abstract

Newborn screening aims to detect rare, inherited metabolic diseases early in newborns. The presymptomatic diagnosis of these diseases, enabling effective therapies, can enhance the quality of life for affected children and their families. However, newborn screening for specific diseases faces challenges, such as false-positive screening results and a need for individualized disease management due to the substantial clinical variability of inherited metabolic diseases. This dissertation aims to develop new mathematical and data-based modeling approaches to support and improve newborn screening. Therefore, data-based machine learning models using data from over two million newborns are developed to enhance diagnostic accuracy in newborn screening for isovaleric aciduria and glutaric aciduria type 1. Furthermore, in the course of this dissertation a genome-based, mathematical whole-body model is developed to depict the metabolism of newborns and infants. This model enables personalized in silico simulations of newborns with an inherited metabolic disease, thereby facilitating individualized disease management. The proposed mathematical and data-based models hold considerable promise for application in personalized medicine and newborn screening, contributing to their improvement and support.