Robust mu-synthesis: Towards a unified glucose control in adults, adolescents and children with T1DM

Abstract

"Type 1 Diabetes Mellitus (T1DM) remains as a severe public health problem in a wide range of population, from children to adults. Recently, the number of diabetics worldwide and morbility rates are increasing. Therefore, emerging technologies as the artificial pancreas (AP) are directing their efforts to improve treatments and to reduce long-term complications. In this work, a cross-age control strategy is proposed to tackle the blood glucose regulation problem in people with T1DM. The contribution of this paper is focused on the blood glucose regulation in T1DM patients, at distinct ages, and it can be controlled in face to physiological uncertain parameters. In other words, a robust model-based controller is proposed via mu-synthesis technique by considering structured uncertainties in physiological meaningful parameters. The proposed controller exhibits the feasibility for blood glucose regulation in virtual diabetic children, adolescents and adults. The relevance of these parameters lies in their high sensitivity to the solutions of a physiological mathematical model; that is, a slight parametric variation can lead to a hyperglycemic scenario. Thus, it is innovative to consider this uncertainties scheme in parameters that are directly related to the glucose dynamics. The robust control algorithm was integrated into the well-known Uva/Padova simulator for T1DM to show the technical viability of this methodology in the available three populations. The outcomes of a control variability grid analysis show that 90.9% of virtual adults are in upper B-zone and 9.09% in B-zone. Likewise for virtual adolescents, 90.9% fall in upper B-zone and 9.09% B-zone. Regarding children, 63.63% lie in upper B-zone, 27.27% B-zone and only 9.09% failure to deal with hypoglycemia. Furthermore, the results are compared to the ones obtained from H-infinity schemes previously reported which also were implemented in the simulator. Despite being a theoretical approach, the results reveal that the proposed cross-age control scheme could be a useful strategy in the development of real PA systems that lead to future clinical trials in humans. (C) 2020 The Franklin Institute. Published by Elsevier Ltd. All rights reserved."