NICU Automatic Oxygen Control with Parameter and Disturbance Estimation

点击翻译按钮获取中文摘要

PROJECT SUMMARY/ABSTRACT The goal of this research is to demonstrate a novel adaptive oxygen control system which will improve the control of oxygen saturation in premature infants who are receiving respiratory support due to underdeveloped lungs. In this work, a clinical study at two sites will be conducted to demonstrate an oxygen control device which is able to continuously adjust oxygen automatically. The study is an equivalence crossover to demonstrate that the device performs at least as well as a trained NICU nurse in limiting fluctuations in FiO2 and maintaining SpO2 within specific parameters prescribed by the treating physician. If there is sufficient evidence, superiority will be investigated. In addition, the study will yield results that will characterize the performance of manual and automatic control alternatives that can be compared. This work is part of an ongoing effort to develop a new technology for controlling oxygen in respiratory support systems for premature infants. The key outcome is clinical data that will show how automatic control of oxygen in premature infants affects the accuracy of control of the oxygen saturation level compared to manual control. The impact of this work is that a novel automatic control system will be developed and tested which will improve the consistency of patient care by automatically adapting the control algorithm to each patient and will lead to a better understanding of the dynamics of the response of neonates to oxygen control. The new patented oxygen control technology uses a parameter estimating extended Kalman filter (PE-EKF) that uses the time history of measurements to estimate the dynamic model parameters of the patients so that the oxygen control system can adapt to a wide range of patient characteristics and conditions. A disturbance estimator also estimates the disturbance level due to unmodeled inputs which cause adverse changes in oxygen saturation. Disturbance estimation allows the control system to quickly quantify the disturbance level and respond by manipulating inspired oxygen to cancel out disturbances that cause desaturation events. The new developments in modeling the system, disturbance estimator, and PE-EKF allow real-time adaptation of the oxygen control system to the changes in the patient during early development and the onset physiological changes. In this work, researchers will be investigating an advanced oxygen control system with a design based on the performance observed during clinical testing and analysis of the system dynamics to further improve the performance.