Kernel-based Estimate of the Insulin Action of Rapid-Acting Insulin in Home-Monitored Data

Summary, in English

To improve glycemic control in insulin-dependent diabetes, estimates of the glucose- lowering effect of insulin-the insulin action (IA)-and the dynamics thereof are essential. The current practice relies on generic estimates of the dynamic shape of the IA and rules-of-thumb to assess the total glucose-lowering effect. In this paper we present individualized estimates based on home-monitored data, incorporating a kernel-based estimate that also accounts for the nonlinear behavior of the IA across the glucose range. The method is successfully validated on 29 patients on continuous insulin pump therapy.

• Author: Fredrik Ståhl; Erik Lindström; Mona Landin-Olsson; Rolf Johansson
• Department/s: ELLIIT: the Linköping-Lund initiative on IT and mobile communication; Department of Automatic Control; Mathematical Statistics; Financial Mathematics Group; Department of Clinical Sciences, Lund; Medicine, Lund; Diabetes lab
• Publishing year: 2015
• Language: English
• Pages: 715-720
• Publication/Series: IFAC-PapersOnLine
• Volume: 48
• Issue: 28
• Links: http://dx.doi.org/10.1016/j.ifacol.2015.12.214
• Document type: Journal article
• Publisher: IFAC Secretariat
• Topic: Medical Biotechnology; Control Engineering
• Status: Published
• Project: DIAdvisor
• Research group: Financial Mathematics Group; Diabetes lab
• ISBN/ISSN/Other: ISSN: 2405-8963