Control of blood glucose is essential for diabetes management. Current digital therapeutic approaches for subjects with Type 1 diabetes mellitus (T1DM) such as the artificial pancreas and insulin bolus calculators leverage machine learning techniques for predicting subcutaneous glucose for improved control. Deep learning has recently been applied in healthcare and medical research to achieve state-of-the-art results in a range of tasks including disease diagnosis, and patient state prediction among others. In this work, we present a deep learning model that is capable of forecasting glucose levels with leading accuracy for simulated patient cases (RMSE = 9.38±0.71 [mg/dL] over a 30-minute horizon, RMSE = 18.87±2.25 [mg/dL] over a 60minute horizon) and real patient cases (RMSE = 21.07±2.35 [mg/dL] for 30-minute, RMSE = 33.27±4.79% for 60-minute).In addition, the model provides competitive performance in providing effective prediction horizon (P H ef f ) with minimal time lag both in a simulated patient dataset (P H ef f = 29.0±0.7 for 30-min and P H ef f = 49.8±2.9 for 60-min) and in a real patient dataset (P H ef f = 19.3±3.1 for 30-min and P H ef f = 29.3±9.4 for 60-min). This approach is evaluated on a dataset of 10 simulated cases generated from the UVa/Padova simulator and a clinical dataset of 10 real cases each containing glucose readings, insulin bolus, and meal (carbohydrate) data. Performance of the recurrent convolutional neural network is benchmarked against four algorithms. The proposed algorithm is implemented on an Android mobile phone, with an execution time of 6ms on a phone compared to an execution time of 780ms on a laptop.

For people with Type 1 diabetes (T1D), forecasting of blood glucose (BG) can be used to effectively avoid hyperglycemia, hypoglycemia and associated complications. The latest continuous glucose monitoring (CGM) technology allows people to observe glucose in real-time. However, an accurate glucose forecast remains a challenge. In this work, we introduce GluNet, a framework that leverages on a personalized deep neural network to predict the probabilistic distribution of short-term (30-60 minutes) future CGM measurements for subjects with T1D based on their historical data including glucose measurements, meal information, insulin doses, and other factors. It adopts the latest deep learning techniques consisting of four components: data pre-processing, label transform/recover, multi-layers of dilated convolution neural network (CNN), and post-processing. The method is evaluated in-silico for both adult and adolescent subjects. The results show significant improvements over existing methods in the literature through a comprehensive comparison in terms of root mean square error (RMSE) (8.88 ± 0.77 mg/dL) with short time lag (0.83 ± 0.40 minutes) for prediction horizons (PH) = 30 mins (minutes), and RMSE (19.90 ± 3.17 mg/dL) with time lag (16.43 ± 4.07 mins) for PH = 60 mins for virtual adult subjects. In addition, GluNet is also tested on two clinical data sets. Results show that it achieves an RMSE (19.28±2.76 mg/dL) with time lag (8.03 ± 4.07 mins) for PH = 30 mins and an RMSE (31.83±3.49 mg/dL) with time lag (17.78±8.00 mins) for PH = 60 mins. These are the best reported results for glucose forecasting when compared with other methods including the neural network for predicting glucose (NNPG), the support vector regression (SVR), the latent variable with exogenous input (LVX), and the auto regression with exogenous input (ARX) algorithm.

(1) Objective: Blood glucose forecasting in type 1 diabetes (T1D) management is a maturing field with numerous algorithms being published and a few of them having reached the commercialisation stage. However, accurate long-term glucose predictions (e.g., >60 min), which are usually needed in applications such as precision insulin dosing (e.g., an artificial pancreas), still remain a challenge. In this paper, we present a novel glucose forecasting algorithm that is well-suited for long-term prediction horizons. The proposed algorithm is currently being used as the core component of a modular safety system for an insulin dose recommender developed within the EU-funded PEPPER (Patient Empowerment through Predictive PERsonalised decision support) project. (2) Methods: The proposed blood glucose forecasting algorithm is based on a compartmental composite model of glucose–insulin dynamics, which uses a deconvolution technique applied to the continuous glucose monitoring (CGM) signal for state estimation. In addition to commonly employed inputs by glucose forecasting methods (i.e., CGM data, insulin, carbohydrates), the proposed algorithm allows the optional input of meal absorption information to enhance prediction accuracy. Clinical data corresponding to 10 adult subjects with T1D were used for evaluation purposes. In addition, in silico data obtained with a modified version of the UVa-Padova simulator was used to further evaluate the impact of accounting for meal absorption information on prediction accuracy. Finally, a comparison with two well-established glucose forecasting algorithms, the autoregressive exogenous (ARX) model and the latent variable-based statistical (LVX) model, was carried out. (3) Results: For prediction horizons beyond 60 min, the performance of the proposed physiological model-based (PM) algorithm is superior to that of the LVX and ARX algorithms. When comparing the performance of PM against the secondly ranked method (ARX) on a 120 min prediction horizon, the percentage improvement on prediction accuracy measured with the root mean square error, A-region of error grid analysis (EGA), and hypoglycaemia prediction calculated by the Matthews correlation coefficient, was 18.8%, 17.9%, and 80.9%, respectively. Although showing a trend towards improvement, the addition of meal absorption information did not provide clinically significant improvements. (4) Conclusion: The proposed glucose forecasting algorithm is potentially well-suited for T1D management applications which require long-term glucose predictions.

This paper presents an algorithm for the computation of full-complexity polytopic robust control invariant (RCI) sets, and the corresponding linear statefeedback control law. The proposed scheme can be applied for linear discretetime systems subject to additive disturbances and structured norm-bounded or polytopic uncertainties. Output, initial condition, and performance constraints are considered. Arbitrary complexity of the invariant polytope is allowed to enable less conservative inner/outer approximations to the RCI sets whereas the RCI set is assumed to be symmetric around the origin. The nonlinearities associated with the computation of such an RCI set structure are overcome through the application of Farkas' theorem and a corollary of the elimination lemma to obtain an initial polytopic RCI set, which is guaranteed to exist under certain conditions. A Newton-like update, which is recursively feasible, is then proposed to yield desirable large/small volume RCI sets. KEYWORDS full-complexity polytope, LMIs, optimization, robust control invariant sets, uncertain systems Int J Robust Nonlinear Control. 2019;29:3587-3605.wileyonlinelibrary.com/journal/rnc

Background: Delivering insulin in type 1 diabetes is a challenging, and potentially risky, activity; hence the importance of including safety measures as part of any insulin dosing or recommender system. This work presents and clinically evaluates a modular safety system that is part of an intelligent insulin dose recommender platform developed within the EU-funded PEPPER project. Methods: The proposed safety system is composed of four modules which use a novel glucose forecasting algorithm. These modules are predictive glucose alerts and alarms; a predictive low-glucose basal insulin suspension module; an advanced rescue carbohydrate recommender for resolving hypoglycemia; and a personalized safety constraint applied to insulin recommendations. The technical feasibility of the proposed safety system was evaluated in a pilot study including eight adult subjects with type 1 diabetes on multiple daily injections over a duration of six weeks. Glycemic control and safety system functioning were compared between the two-weeks run-in period and the end point at eight weeks. A standard insulin bolus calculator was employed to recommend insulin doses. Results: Overall, glycemic control improved over the evaluated period. In particular, percentage time in the hypoglycemia range (<3.0 mmol/l) significantly decreased from 0.82% (0.05-4.79) at run-in to 0.33% (0.00-0.93) at endpoint ( P = .02). This was associated with a significant increase in percentage time in target range (3.9-10.0 mmol/l) from 52.8% (38.3-61.5) to 61.3% (47.5-71.7) ( P = .03). There was also a reduction in number of carbohydrate recommendations. Conclusion: A safety system for an insulin dose recommender has been proven to be a viable solution to reduce the number of adverse events associated to glucose control in type 1 diabetes.