A grid‐based tool for optimal performance monitoring of a glycemic regulator. (3rd December 2015)

• Record Type: Journal Article
• Authors: Ávila, Luis; Martínez, Ernesto
• Is Part Of: Optimal control applications and methods. Volume 37:Number 6(2016:Nov./Dec.)
• Page Start: 1103
• Page End: 1121
• Publication Date: 2015-12-03
• Subjects: artificial pancreas -- optimal control applications -- performance monitoring -- stochastic optimal control -- type 1 diabetes -- uncertain systems; Control theory -- Periodicals; Mathematical optimization -- Periodicals; 629.8312
• DOI: 10.1002/oca.2220
• Languages: English
• ISSNs: 0143-2087

Abstract

Summary: Recent technology breakthroughs towards a fully automated artificial pancreas give rise to the need of new monitoring tools aiming at increasing both reliability and performance of a closed‐loop glycemic regulator. Based on error grid analysis, an insightful monitoring tool is proposed to assess if a given closed‐loop implementation respects its specification of an optimally performing glycemic regulator under uncertainty. The optimal behavior specification is obtained using linearly solvable Markov decision processes, whereby the Bellman optimality equation is made linear through an exponential transformation that allows obtaining the optimal control policy in an explicit form. The specification for the desired glucose dynamics is learned using Gaussian processes for state transitions in an optimally performing artificial pancreas. By means of the proposed grid, the specification is vis‐à‐vis compared with glucose sensor readings so that any significant deviation from the expected closed‐loop performance under abnormal or faulty scenarios can be detected. Copyright © 2015 John Wiley & Sons, Ltd.