Adaptive system identification for estimating future glucose concentrations and hypoglycemia alarms

Abstract

Many patients with diabetes experience high variability in glucose concentrations that includes prolonged hyperglycemia or hypoglycemia. Models predicting a subject’s future glucose concentrations can be used for preventing such conditions by providing early alarms. This paper presents a time-series model that captures dynamical changes in the glucose metabolism. Adaptive system identification is proposed to estimate model parameters which enable the adaptation of the model to inter-/intra-subject variation and glycemic disturbances. It consists of online parameter identification using the weighted recursive least squares method and a change detection strategy that monitors variation in model parameters. Univariate models developed from a subject’s continuous glucose measurements are compared to multivariate models that are enhanced with continuous metabolic, physical activity and lifestyle information from a multi-sensor body monitor. A real life application for the proposed algorithm is demonstrated on early (30 min in advance) hypoglycemia detection.

Physical Activity Measured by Physical Activity Monitoring System Correlates with Glucose Trends Reconstructed from Continuous Glucose Monitoring

“…Of course, the effective quantitative incorporation of PA information within glucose predictors and closed-loop systems will require in-depth future investigations. For instance, with regard to prediction algorithms, PA information could be exploited to dynamically modulate the forgetting factor typically used in low-order time-varying autoregressive/polynomial models (e.g., the forgetting factor of simple time-varying prediction algorithms as those proposed by Eren-Oruklu et al 23,53 and Sparacino et al 54 could be decreased in real time in the presence of PA).…”

“…[14][15][16][17][18] However, to be effective, all these CGM-based applications should be also able to describe changes in glucose dynamics caused by perturbing factors, such as meals, insulin injections, exercise, stress, etc. The relationship between these variations and meals/insulin information has been widely investigated in the literature, 19 stimulating their inclusion in short-time glucose prediction algorithms [20][21][22][23] and in control algorithms. 24,25 Nevertheless, inclusion of PA information is still challenging because dealing with PA is, to the best of our knowledge, less explored, and models describing the effects of PA on glucose dynamics are mainly qualitative or validated on small datasets.…”

Physical Activity Capture Technology With Potential for Incorporation Into Closed-Loop Control for Type 1 Diabetes

“…Significant efforts have been made to model the aforementioned parameters and understand their effect on glucose-insulin dynamics by incorporating/adapting them into CLC simulators. 6,7 Even though the day to day effect of physical activity on glucose status poses major challenges for T1D, limited research in this area has been conducted. 8 Since obesity and type 2 diabetes mellitus (T2D), 2 major disorders of whole body energy balance, continue to rise worldwide, most research to-date regarding physical activity capture device use has been pursued in these 2 disorders.…”

How Much Is Short-Term Glucose Prediction in Type 1 Diabetes Improved by Adding Insulin Delivery and Meal Content Information to CGM Data? A Proof-of-Concept Study

“…-autoregressive with exogenous input models exploiting information on CHO and insulin therapy 11,12 -autoregressive with moving average with exogenous inputs models accounting for food intake, physical activity, emotional stimuli, and lifestyle; 13 physical activity and insulin on board information; 14 and insulin and CHO information 15 -latent variable-based predictors 16 -random forests, support vector-based algorithms, and…”