Personalized Feeding Infusion Rate of Nutrition to Improve Glycemic Variability Control in Critical Care Patients

Abstract

Control-based algorithms in the intensive care unit (ICU) patients have been developed to deliver a sufficient amount of insulin, but optimizing the rate of feeding of nutrition in ICU patients to improve glycemic variability control has not been done yet. Continuous feeding is commonly used for nutrition in critically ill patients who cannot be fed orally to maintain a normal blood sugar concentration, but optimizing its rate, for these individuals, is needed to avoid the adverse outcomes caused by medications such as insulin. This paper develops a control-based algorithm combines a predictive control algorithm with a revised nonlinear compartmental model used in the ICU to design personalized feeding function rates to improve patient glycemic variability. Our control algorithm is robust and acts very quickly to avoid medical intervention effects.