Data-driven decisions in neonatology and pediatric surgery
Data-driven decisions in neonatology and pediatric surgery
Artificial intelligence assists with complex decisions for vulnerable babies
Extremely premature babies are highly vulnerable. One of the greatest dangers is necrotizing enterocolitis (NEC), a bowel disease that can strike unexpectedly. Within a short time, a baby can become seriously ill and may require surgery.
This dissertation of Rosa Verhoeven investigates how artificial intelligence (AI) can support doctors and parents in cases of NEC. It shows that AI can provide early warnings for this disease based on standard measurements including heart rate and oxygen levels. This allows doctors to be better prepared and to intervene more quickly.
AI can also assist with ethical decisions. Sometimes surgery is necessary to save the life of a baby with NEC, but such an operation is invasive, the chances of survival are uncertain, and survivors often have a reduced quality of life. In some cases, it may be better to ensure that the baby can pass away as calmly and painlessly as possible.
Using AI, we gained insight into how doctors make these decisions in the Netherlands. In this dissertation, we compared whether AI predictions align with real-world choices. We also developed an AI model to compare decision-making among doctors in Europe.
However, AI does not always provide a solution. Using other methods, we investigated how parents make these decisions. These insights help healthcare providers support parents more effectively and make conversations more personal.
This research demonstrates how technology can contribute to better care, informed communication between doctors and parents, and greater understanding in emotionally challenging situations.