Adaptive Personalized Prior-Knowledge-Informed Model Predictive Control for Type 1 Diabetes.

This work considers the problem of adaptive prior-informed model predictive control (MPC) formulations that explicitly incorporate prior knowledge in the model development and is robust to missing data in the output measurements. The proposed prediction model is based on a latent variables model to extract glycemic dynamics from highly-correlated data and incorporates prior knowledge of exponential stability to improve the prediction ability. Missing data structures are formulated to enable model predictions when output measurements are missing for short periods of time. Based on the latent variables model, the MPC strategy and adaptive rules are developed to automatically tune the aggressiveness of the MPC. The adaptive prior-knowledge-informed MPC is evaluated with computer simulations for the control of blood glucose concentrations in people with Type 1 diabetes (T1D) using simulated virtual patients. Due to the variability among people with T1D, the hyperparameters of the prior-knowledge-informed model are personalized to individual subjects. The percentage of time spent in the target range is 76.48% when there are no missing data and 76.52% when there are missing data episodes lasting up to 30 mins (6 samples). Incorporating the adaptive rules further improves the percentage of time in target range to 84.58% and 84.88% for cases with no missing data and missing data, respectively. The proposed adaptive prior-informed MPC formulation provides robust, effective, and safe regulation of glucose concentration in T1D despite disturbances and missing measurements.

Sleep quality and glycaemic variability in a real-life setting in adults with type 1 diabetes.

AIMS/HYPOTHESIS: Suboptimal subjective sleep quality is very common in adults with type 1 diabetes. Reducing glycaemic variability is a key objective in this population. To date, no prior studies have examined the associations between objectively measured sleep quality and overnight glycaemic variability in adults with type 1 diabetes. We aimed to test the hypothesis that poor sleep quality would be associated with greater overnight glycaemic variability. METHODS: Data were collected in the home setting from 20 adults (ten male and ten female participants) with type 1 diabetes who were current insulin pump users. Simultaneous assessments of objective sleep quality (Zmachine Insight+) and continuous glucose monitoring (CGM) were performed over multiple nights (up to 15 nights) in each participant. Due to the real-life nature of this study, the participants kept their usual CGM alerts for out-of-range glucose values. Sleep quality was categorised as 'good' or 'poor' using a composite of objective sleep features (i.e. sleep efficiency, wake after sleep onset and number of awakenings) based on the National Sleep Foundation's consensus criteria. Glycaemic variability was quantified using SD and CV of overnight glucose values based on overnight CGM profiles. RESULTS: A total of 170 nights were analysed. Overall, 86 (51%) nights were categorised as good sleep quality, and 84 (49%) nights were categorised as poor sleep quality. Using linear mixed-effects models, poor sleep quality was significantly associated with greater glycaemic variability as quantified by SD (coefficient: 0.39 [95% CI 0.10, 0.67], p = 0.009) and CV (coefficient: 4.35 [95% CI 0.8, 7.9], p = 0.02) of overnight glucose values, after accounting for age, sex, BMI and overnight insulin dose. There was large inter- and intra-individual variability in sleep and glycaemic characteristics. Both pooled and individual-level data revealed that the nights with poor sleep quality had larger SDs and CVs, and, conversely, the nights with good sleep quality had smaller SDs and CVs. No associations were found between sleep quality and time spent in the target glucose range, or above or below the target glucose range, where CGM alarms are most likely to occur. CONCLUSIONS/INTERPRETATION: Objectively measured sleep quality is associated with overnight glycaemic variability in adults with type 1 diabetes. These findings highlight an important role of sleep quality in overnight glycaemic control in type 1 diabetes. They also provide a strong incentive to both patients and healthcare providers for considering sleep quality in personalised type 1 diabetes glycaemic management plans. Future studies should investigate the mechanisms linking sleep quality to glycaemic variability.

Prior Informed Regularization of Recursively Updated Latent-Variables-Based Models with Missing Observations.

Many data-driven modeling techniques identify locally valid, linear representations of time-varying or nonlinear systems, and thus the model parameters must be adaptively updated as the operating conditions of the system vary, though the model identification typically does not consider prior knowledge. In this work, we propose a new regularized partial least squares (rPLS) algorithm that incorporates prior knowledge in the model identification and can handle missing data in the independent covariates. This latent variable (LV) based modeling technique consists of three steps. First, a LV-based model is developed on the historical time series data. In the second step, the missing observations in the new incomplete data sample are estimated. Finally, the future values of the outputs are predicted as a linear combination of estimated scores and loadings. The model is recursively updated as new data are obtained from the system. The performance of the proposed rPLS and rPLS with exogenous inputs (rPLSX) algorithms are evaluated by modeling variations in glucose concentration (GC) of people with Type 1 diabetes (T1D) in response to meals and physical activities for prediction windows up to one hour, or 12 sampling instances, into the future. The proposed rPLS family of GC prediction models are evaluated with both in-silico and clinical experiment data and compared with the performance of recursive time series and kernel-based models. The root mean squared error (RMSE) with simulated subjects in the multivariable T1D simulator where physical activity effects are incorporated in GC variations are 2.52 and 5.81 mg/dL for 30 and 60 mins ahead predictions (respectively) when information for all meals and physical activities are used, increasing to 2.70 and 6.54 mg/dL (respectively) when meals and activities occurred, but the information is with-held from the modeling algorithms. The RMSE is 10.45 and 14.48 mg/dL for clinical study with prediction horizons of 30 and 60 mins, respectively. The low RMSE values demonstrate the effectiveness of the proposed rPLS approach compared to the conventional recursive modeling algorithms.

Determining Physical Activity Characteristics from Wristband Data for Use in Automated Insulin Delivery Systems.

Algorithms that can determine the type of physical activity (PA) and quantify the intensity can allow precision medicine approaches, such as automated insulin delivery systems that modulate insulin administration in response to PA. In this work, data from a multi-sensor wristband is used to design classifiers to distinguish among five different physical states (PS) (resting, activities of daily living, running, biking, and resistance training), and to develop models to estimate the energy expenditure (EE) of the PA for diabetes therapy. The data collected are filtered, features are extracted from the reconciled signals, and the extracted features are used by machine learning algorithms, including deep-learning techniques, to obtain accurate PS classification and EE estimation. The various machine learning techniques have different success rates ranging from 75.7% to 94.8% in classifying the five different PS. The deep neural network model with long short-term memory has 94.8% classification accuracy. We achieved 0.5 MET (Metabolic Equivalent of Task) root-mean-square error for EE estimation accuracy, relative to indirect calorimetry with randomly selected testing data (10% of collected data). We also demonstrate a 5% improvement in PS classification accuracy and a 0.34 MET decrease in the mean absolute error when using multi-sensor approach relative to using only accelerometer data.

Online Glucose Prediction Using Computationally Efficient Sparse Kernel Filtering Algorithms in Type-1 Diabetes.

Streaming data from continuous glucose monitoring (CGM) systems enable the recursive identification of models to improve estimation accuracy for effective predictive glycemic control in patients with type-1 diabetes. A drawback of conventional recursive identification techniques is the increase in computational requirements, which is a concern for online and real-time applications such as the artificial pancreas systems implemented on handheld devices and smartphones where computational resources and memory are limited. To improve predictions in such computationally constrained hardware settings, efficient adaptive kernel filtering algorithms are developed in this paper to characterize the nonlinear glycemic variability by employing a sparsification criterion based on the information theory to reduce the computation time and complexity of the kernel filters without adversely deteriorating the predictive performance. Furthermore, the adaptive kernel filtering algorithms are designed to be insensitive to abnormal CGM measurements, thus compensating for measurement noise and disturbances. As such, the sparsification-based real-time model update framework can adapt the prediction models to accurately characterize the time-varying and nonlinear dynamics of glycemic measurements. The proposed recursive kernel filtering algorithms leveraging sparsity for improved computational efficiency are applied to both in-silico and clinical subjects, and the results demonstrate the effectiveness of the proposed methods.

Automated Insulin Delivery Algorithms.

IN BRIEF Automated insulin delivery (AID; also known as artificial pancreas) has improved the regulation of blood glucose concentrations, reduced the frequency of hyperglycemic and hypoglycemic episodes, and improved the quality of life of people with diabetes and their families. Three different types of algorithms-proportional-integral-derivative control, model predictive control, and fuzzy-logic knowledge-based systems-have been used in AID control systems. This article will highlight the foundations of these algorithms and discuss their strengths and limitations. Multivariable artificial pancreas and dual-hormone (insulin and glucagon) systems will be introduced.

Simulation Software for Assessment of Nonlinear and Adaptive Multivariable Control Algorithms: Glucose - Insulin Dynamics in Type 1 Diabetes.

A simulator for testing automatic control algorithms for nonlinear systems with time-varying parameters, variable time delays, and uncertainties is developed. It is based on simulation of virtual patients with Type 1 diabetes (T1D). Nonlinear models are developed to describe glucose concentration (GC) variations based on user-defined scenarios for meal consumption, insulin administration, and physical activity. They compute GC values and physiological variables, such as heart rate, skin temperature, accelerometer, and energy expenditure, that are indicative of physical activities affecting GC dynamics. This is the first simulator designed for assessment of multivariable controllers that consider supplemental physiological variables in addition to GC measurements to improve glycemic control. Virtual patients are generated from distributions of identified model parameters using clinical data. The simulator will enable testing and evaluation of new control algorithms proposed for automated insulin delivery as well as various control algorithms for nonlinear systems with uncertainties, time-varying parameters and delays.

Plasma-Insulin-Cognizant Adaptive Model Predictive Control for Artificial Pancreas Systems.

An adaptive model predictive control (MPC) algorithm with dynamic adjustments of constraints and objective function weights based on estimates of the plasma insulin concentration (PIC) is proposed for artificial pancreas (AP) systems. A personalized compartment model that translates the infused insulin into estimates of PIC is integrated with a recursive subspace-based system identification to characterize the transient dynamics of glycemic measurements. The system identification approach is able to identify stable, reliable linear time-varying models from closed-loop data. An MPC algorithm using the adaptive models is designed to compute the optimal exogenous insulin delivery for AP systems without requiring any manually-entered meal information. A dynamic safety constraint derived from the estimation of PIC is incorporated in the adaptive MPC to improve the efficacy of the AP and prevent insulin overdosing. Simulation case studies demonstrate the performance of the proposed adaptive MPC algorithm.

Multi-level Supervision and Modification of Artificial Pancreas Control System.

Artificial pancreas (AP) systems provide automated regulation of blood glucose concentration (BGC) for people with type 1 diabetes (T1D). An AP includes three components: a continuous glucose monitoring (CGM) sensor, a controller calculating insulin infusion rate based on the CGM signal, and a pump delivering the insulin amount calculated by the controller to the patient. The performance of the AP system depends on successful operation of these three components. Many APs use model predictive controllers that rely on models to predict BGC and to calculate the optimal insulin infusion rate. The performance of model-based controllers depends on the accuracy of the models that is affected by large dynamic changes in glucose-insulin metabolism or equipment performance that may move the operating conditions away from those used in developing the models and designing the control system. Sensor errors and missing signals will cause calculation of erroneous insulin infusion rates. And the performance of the controller may vary at each sampling step and each period (meal, exercise, and sleep), and from day to day. Here we describe a multi-level supervision and controller modification (ML-SCM) module is developed to supervise the performance of the AP system and retune the controller. It supervises AP performance in 3 time windows: sample level, period level, and day level. At sample level, an online controller performance assessment sub-module will generate controller performance indexes to evaluate various components of the AP system and conservatively modify the controller. A sensor error detection and signal reconciliation module will detect sensor error and reconcile the CGM sensor signal at each sample. At period level, the controller performance is evaluated with information collected during a certain time period and the controller is tuned more aggressively. At the day level, the daily CGM ranges are further analyzed to determine the adjustable range of controller parameters used for sample level and period level. Thirty subjects in the UVa/Padova metabolic simulator were used to evaluate the performance of the ML-SCM module and one clinical experiment is used to illustrate its performance in a clinical environment. The results indicate that the AP system with an ML-SCM module has a safer range of glucose concentration distribution and more appropriate insulin infusion rate suggestions than an AP system without the ML-SCM module.

Model-Fusion-Based Online Glucose Concentration Predictions in People with Type 1 Diabetes.

Accurate predictions of glucose concentrations are necessary to develop an artificial pancreas (AP) system for people with type 1 diabetes (T1D). In this work, a novel glucose forecasting paradigm based on a model fusion strategy is developed to accurately characterize the variability and transient dynamics of glycemic measurements. To this end, four different adaptive filters and a fusion mechanism are proposed for use in the online prediction of future glucose trajectories. The filter fusion mechanism is developed based on various prediction performance indexes to guide the overall output of the forecasting paradigm. The efficiency of the proposed model fusion based forecasting method is evaluated using simulated and clinical datasets, and the results demonstrate the capability and prediction accuracy of the data-based fusion filters, especially in the case of limited data availability. The model fusion framework may be used in the development of an AP system for glucose regulation in patients with T1D.

Multivariable Adaptive Artificial Pancreas System in Type 1 Diabetes.

PURPOSE OF REVIEW: The review summarizes the current state of the artificial pancreas (AP) systems and introduces various new modules that should be included in future AP systems. RECENT FINDINGS: A fully automated AP must be able to detect and mitigate the effects of meals, exercise, stress and sleep on blood glucose concentrations. This can only be achieved by using a multivariable approach that leverages information from wearable devices that provide real-time streaming data about various physiological variables that indicate imminent changes in blood glucose concentrations caused by meals, exercise, stress and sleep. The development of a fully automated AP will necessitate the design of multivariable and adaptive systems that use information from wearable devices in addition to glucose sensors and modify the models used in their model-predictive alarm and control systems to adapt to the changes in the metabolic state of the user. These AP systems will also integrate modules for controller performance assessment, fault detection and diagnosis, machine learning and classification to interpret various signals and achieve fault-tolerant control. Advances in wearable devices, computational power, and safe and secure communications are enabling the development of fully automated multivariable AP systems.

Use of Wearable Sensors and Biometric Variables in an Artificial Pancreas System.

An artificial pancreas (AP) computes the optimal insulin dose to be infused through an insulin pump in people with Type 1 Diabetes (T1D) based on information received from a continuous glucose monitoring (CGM) sensor. It has been recognized that exercise is a major challenge in the development of an AP system. The use of biometric physiological variables in an AP system may be beneficial for prevention of exercise-induced challenges and better glucose regulation. The goal of the present study is to find a correlation between biometric variables such as heart rate (HR), heat flux (HF), skin temperature (ST), near-body temperature (NBT), galvanic skin response (GSR), and energy expenditure (EE), 2D acceleration-mean of absolute difference (MAD) and changes in glucose concentrations during exercise via partial least squares (PLS) regression and variable importance in projection (VIP) in order to determine which variables would be most useful to include in a future artificial pancreas. PLS and VIP analyses were performed on data sets that included seven different types of exercises. Data were collected from 26 clinical experiments. Clinical results indicate ST to be the most consistently important (important for six out of seven tested exercises) variable over all different exercises tested. EE and HR are also found to be important variables over several types of exercise. We also found that the importance of GSR and NBT observed in our experiments might be related to stress and the effect of changes in environmental temperature on glucose concentrations. The use of the biometric measurements in an AP system may provide better control of glucose concentration.

Hybrid online sensor error detection and functional redundancy for systems with time-varying parameters.

Supervision and control systems rely on signals from sensors to receive information to monitor the operation of a system and adjust manipulated variables to achieve the control objective. However, sensor performance is often limited by their working conditions and sensors may also be subjected to interference by other devices. Many different types of sensor errors such as outliers, missing values, drifts and corruption with noise may occur during process operation. A hybrid online sensor error detection and functional redundancy system is developed to detect errors in online signals, and replace erroneous or missing values detected with model-based estimates. The proposed hybrid system relies on two techniques, an outlier-robust Kalman filter (ORKF) and a locally-weighted partial least squares (LW-PLS) regression model, which leverage the advantages of automatic measurement error elimination with ORKF and data-driven prediction with LW-PLS. The system includes a nominal angle analysis (NAA) method to distinguish between signal faults and large changes in sensor values caused by real dynamic changes in process operation. The performance of the system is illustrated with clinical data continuous glucose monitoring (CGM) sensors from people with type 1 diabetes. More than 50,000 CGM sensor errors were added to original CGM signals from 25 clinical experiments, then the performance of error detection and functional redundancy algorithms were analyzed. The results indicate that the proposed system can successfully detect most of the erroneous signals and substitute them with reasonable estimated values computed by functional redundancy system.

Artificial Intelligence and Machine Learning for Improving Glycemic Control in Diabetes: Best Practices, Pitfalls, and Opportunities.

OBJECTIVE: Artificial intelligence and machine learning are transforming many fields including medicine. In diabetes, robust biosensing technologies and automated insulin delivery therapies have created a substantial opportunity to improve health. While the number of manuscripts addressing the topic of applying machine learning to diabetes has grown in recent years, there has been a lack of consistency in the methods, metrics, and data used to train and evaluate these algorithms. This manuscript provides consensus guidelines for machine learning practitioners in the field of diabetes, including best practice recommended approaches and warnings about pitfalls to avoid. METHODS: Algorithmic approaches are reviewed and benefits of different algorithms are discussed including importance of clinical accuracy, explainability, interpretability, and personalization. We review the most common features used in machine learning applications in diabetes glucose control and provide an open-source library of functions for calculating features, as well as a framework for specifying data sets using data sheets. A review of current data sets available for training algorithms is provided as well as an online repository of data sources. SIGNIFICANCE: These consensus guidelines are designed to improve performance and translatability of new machine learning algorithms developed in the field of diabetes for engineers and data scientists.

Detection of Meals and Physical Activity Events From Free-Living Data of People With Diabetes.

BACKGROUND: Predicting carbohydrate intake and physical activity in people with diabetes is crucial for improving blood glucose concentration regulation. Patterns of individual behavior can be detected from historical free-living data to predict meal and exercise times. Data collected in free-living may have missing values and forgotten manual entries. While machine learning (ML) can capture meal and exercise times, missing values, noise, and errors in data can reduce the accuracy of ML algorithms. METHODS: Two recurrent neural networks (RNNs) are developed with original and imputed data sets to assess detection accuracy of meal and exercise events. Continuous glucose monitoring (CGM) data, insulin infused from pump data, and manual meal and exercise entries from free-living data are used to predict meals, exercise, and their concurrent occurrence. They contain missing values of various lengths in time, noise, and outliers. RESULTS: The accuracy of RNN models range from 89.9% to 95.7% for identifying the state of event (meal, exercise, both, or neither) for various users. "No meal or exercise" state is determined with 94.58% accuracy by using the best RNN (long short-term memory [LSTM] with 1D Convolution). Detection accuracy with this RNN is 98.05% for meals, 93.42% for exercise, and 55.56% for concurrent meal-exercise events. CONCLUSIONS: The meal and exercise times detected by the RNN models can be used to warn people for entering meal and exercise information to hybrid closed-loop automated insulin delivery systems. Reliable accuracy for event detection necessitates powerful ML and large data sets. The use of additional sensors and algorithms for detecting these events and their characteristics provides a more accurate alternative.

Multivariable Automated Insulin Delivery System for Handling Planned and Spontaneous Physical Activities.

BACKGROUND: Hybrid closed-loop control of glucose levels in people with type 1 diabetes mellitus (T1D) is limited by the requirements on users to manually announce physical activity (PA) and meals to the artificial pancreas system. Multivariable automated insulin delivery (mvAID) systems that can handle unannounced PAs and meals without any manual announcements by the user can improve glycemic control by modulating insulin dosing in response to the occurrence and intensity of spontaneous physical activities. METHODS: An mvAID system is developed to supplement the glucose measurements with additional physiological signals from a wristband device, with the signals analyzed using artificial intelligence algorithms to automatically detect the occurrence of PA and estimate its intensity. This additional information gained from the physiological signals enables more proactive insulin dosing adjustments in response to both planned exercise and spontaneous unanticipated physical activities. RESULTS: In silico studies of the mvAID illustrate the safety and efficacy of the system. The mvAID is translated to pilot clinical studies to assess its performance, and the clinical experiments demonstrate an increased time in range and reduced risk of hypoglycemia following unannounced PA and meals. CONCLUSIONS: The mvAID systems can increase the safety and efficacy of insulin delivery in the presence of unannounced physical activities and meals, leading to improved lives and less burden on people with T1D.

Adaptive System Identification for Estimating Future Glucose Concentrations and Hypoglycemia Alarms.

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.

Meal and Physical Activity Detection from Free-living Data for Discovering Disturbance Patterns to Glucose Levels in People with Diabetes.

Objective: Interpretation of time series data collected in free-living has gained importance in chronic disease management. Some data are collected objectively from sensors and some are estimated and entered by the individual. In type 1 diabetes (T1D), blood glucose concentration (BGC) data measured by continuous glucose monitoring (CGM) systems and insulin doses administered can be used to detect the occurrences of meals and physical activities and generate the personal daily living patterns for use in automated insulin delivery (AID). Methods: Two challenges in time-series data collected in daily living are addressed: data quality improvement and detection of unannounced disturbances to BGC. CGM data have missing values for varying periods of time and outliers. People may neglect reporting their meal and physical activity information. In this work, novel methods for preprocessing real-world data collected from people with T1D and detection of meal and exercise events are presented. Four recurrent neural network (RNN) models are investigated to detect the occurrences of meals and physical activities disjointly or concurrently. Results: RNNs with long short-term memory (LSTM) with 1D convolution layers and bidirectional LSTM with 1D convolution layers have average accuracy scores of 92.32% and 92.29%, and outper-form other RNN models. The F1 scores for each individual range from 96.06% to 91.41% for these two RNNs. Conclusions: RNNs with LSTM and 1D convolution layers and bidirectional LSTM with 1D convolution layers provide accurate personalized information about the daily routines of individuals. Significance: Capturing daily behavior patterns enables more accurate future BGC predictions in AID systems and improves BGC regulation.

Observational Study of Glycemic Impact of Anticipatory and Early-Race Athletic Competition Stress in Type 1 Diabetes.

Athletic competitions and the associated psychological stress are a challenge for people with type 1 diabetes (T1D). This study aims to understand the influence of anticipatory and early race competition stress on blood glucose concentrations and to identify personality, demographic, or behavioral traits indicative in the scope of the impact. Ten recreational athletes with T1D competed in an athletic competition and an exercise-intensity matched non-competition "training" session for comparison. The two hours prior to exercise and the first 30 minutes of exercise were compared between the paired exercise sessions to assess the influence of anticipatory and early race stress. The effectiveness index, average CGM glucose, and the ingested carbohydrate to injected insulin ratio were compared between the paired sessions through regression. In 9 of 12 races studied, an elevated CGM for the race over the individual training session was observed. The rate of change of the CGM during the first 30 minutes of exercise notably differed between the race and training (p = 0.02) with a less rapid decline in CGM occurring during the race for 11 of 12 paired sessions and an increasing CGM trend during the race for 7 of the 12 sessions with the rate of change (mean ± standard deviation) as 1.36 ± 6.07 and -2.59 ± 2.68 mg/dL per 5 minutes for the race and training, respectively. Individuals with longer durations of diabetes often decreased their carbohydrate-to-insulin ratio on race day, taking more insulin, than on the training day while the reverse was noted for those newly diagnosed (r = -0.52, p = 0.05). The presence of athletic competition stress can impact glycemia. With an increasing duration of diabetes, the athletes may be expecting elevated competition glucose concentrations and take preventive measures.