Modeling and monitoring cerebral blood flow in premature infants with patent ductus arteriosus using ultrasound Doppler technique

Abstract

The aim of this thesis was to simulate cerebral bloodflow of premature infants with patent ductus arteriosus, using a lumped model. A linear Windkessel-3 model with constant peripheral resistance, and a non-linear Windkessel-3 model with pressure-dependent peripheral resistance was constructed. The models was constructed with the Simscape add-on tool in MATLAB. Velocity measurements from the finger was done during a physiological test, where the arm was starting from a stretched-out ordinary position and moved to a vertically upraised position. This test was simulated using the models and the simulation results from both models was compared to see which model that produced best simulation results. The simulation results of the non-linear model gave the best fit, and similar values of the change in peripheral resistance and pulsatility index, compared to the measurements. The models was then modified to a linear and a non-linear lumped model, that represents the arterial system from the heart to the intracranial arteries. The lumped models was used to simulate the cerebral blood flow of an infant with patent ductus arteriosus. The results from the models gave similarly results and good correspondence compared to the in vivo results.