Keywords: Congenital diaphragmatic Hernia, Artificial Intelligence, Outcome prediction, Imaging
 

"Outcome prediction in newborns with congenital diaphragmatic hernia: From imaging to artificial intelligence Imag(in)ing the future of CDH"

Congenital diaphragmatic hernia (CDH) results from abnormal fetal development of the diaphragm and migration of abdominal organs into the thorax through the diaphragmatic defect. Herniation impairs lung development, causing pulmonary hypoplasia and persistent pulmonary hypertension (PH). The combined assessment of lung size, liver position, and defect side by prenatal imaging allows the stratification of CDH fetuses into different groups correlating with perinatal mortality and morbidity. Recently, methods based on Artificial Intelligence (AI) have been developed to support the analysis of medical data in the neonatal field, but not in CDH neonates yet. This thesis investigates the key contribution of perinatal imaging in conjunction with clinical features in predicting mortality and morbidity in CDH neonates, with a particular focus on the emerging role of fetal MRI and AI. This thesis supports the application of machine learning (ML) and deep learning (DL) approaches for the development of predictive algorithms for postnatal outcomes, especially for severe PH and survival. AI in CDH will contribute to accurately defining prognosis, guiding early targeted interventions and personalized management, and improving the overall quality of care.

Click here for the live stream.