State estimation and predictive control applied to the treatment of the hypoxic-ischemic encephalopathy in neonates
Helcio R. B. Orlande (Programa de Engenharia Mecânica, COPPE-UFRJ)

The hypoxic-ischemic encephalopathy in neonates is a neurological disorder characterized by the lack of oxygen (hypoxia) and low blood flow (ischemia), often related to perinatal asphyxia due to complications during birth. The main treatment for hypoxic-ischemic encephalopathy in neonates is the cooling of affected regions. Systemic cooling can be achieved by wrapping the body of the neonate with a blanket containing channels through which cold water is circulated. Alternatively, local cooling can be performed by surrounding the head of the neonate with a cap, in which cold water flows through channels, while the remaining body can be warmed by a radiator in the incubator. This seminar summarizes the works recently performed by the authors on the solution of inverse problems and predictive control related to the hypothermia treatment of the hypoxic-ischemic encephalopathy in neonates. The inverse problem involved the estimation of the brain temperature from the information provided by other temperature measurements available during the treatment. The inverse problem was solved as a state estimation problem with the Sampling Importance Resampling (SIR) algorithm of the particle filter method. The solution of the inverse problem was verified with simulated measurements, and ultimately validated with actual experimental data obtained during the local cooling treatment of a neonate in a pediatric intensive care unit. The combined application of the particle filter method and stochastic model predictive control was also numerically examined, in order to observe and control the body temperatures during the cooling treatment of a neonate.