Search

Impaired glucagon secretion is a major component of glucose intolerance in type 2 diabetes mellitus (T2D). Glucagon secretion exhibits heterogenous patterns in individuals and across glucose tolerance diagnoses. Characterization of the range of glucagon secretion patterns can help clinicians personalize diabetes care based on glucagon characteristics in addition to glucose and insulin profiles. A total of 102 subjects with normal glucose tolerance, impaired glucose tolerance, and T2D had their glucagon profiles recorded in response to an oral glucose tolerance test. Shapelet analysis was used to identify the most descriptive patterns of early glucagon secretion, and spectral biclustering was employed to identify biclusters of associated subjects and shapelets. The dynamics of glucose, insulin, and glucagon secretion in each cluster were evaluated to identify overall patterns, and the characteristics of the subjects in each cluster were compared. Three clusters were chosen to represent the glucagon patterns. Membership in these three clusters was interpreted based on the presence or lack of extrema in the first 30 min after oral carbohydrate intake. Cluster 1 (n = 23) had a minimum at 30 min and only negative trends. Cluster 2 had a minimum at 10 min and a maximum at 20 min (n = 25). Cluster 3 (n = 40) had a maximum at 10 min and a minimum at 20 min. Subjects in cluster 1 had the lowest average fasting plasma glucose (90.17 mg/dL) and average age (41.39 years) and the highest HOMA-beta score (87.5%), while subjects in cluster 2 had the highest average fasting plasma glucose (102.56 mg/dL) and average age (53.16 years) and the lowest HOMA-beta score (55.77%). Characterization of glucagon dynamics, in addition to glucose and insulin, can aid in personalized treatment approaches and provide greater insight about the underlying dysfunction in glucose regulation.

In this article, Phillips-type Bernstein operators ( ℬ m , q t F ) ( t , s ) ({{\mathcal{ {\mathcal B} }}}_{m,q}^{t}F)\left(t,s) and ( ℬ n , q s F ) ( t , s ) ({{\mathcal{ {\mathcal B} }}}_{n,q}^{s}F)\left(t,s) , their products, and Boolean sum based on q-integer have been studied on a triangle with all curved sides. Furthermore, convergence of iterates of these operators have been analyzed using the weakly Picard operators technique and the contraction principle.

Background: Adaptive model predictive control (MPC) algorithms that recursively update the glucose prediction model are shown to be promising in the development of fully automated multivariable artificial pancreas systems. However, the recursively updated glycemic prediction models do not explicitly consider prior knowledge in the identification of the model parameters. Prior information of the glycemic effects of meals and physical activity can improve model accuracy and yield better glycemic control algorithms. Methods: A glucose prediction model based on regularized partial least squares (rPLS) method where the prior information is encoded as the regularization term is developed to provide accurate predictions of the future glucose concentrations. An adaptive MPC is developed that incorporates dynamic trajectories for the glucose setpoint and insulin dosing constraints based on the estimated plasma insulin concentration (PIC). The proposed adaptive MPC algorithm is robust to disturbances caused by unannounced meals and physical activities even in cases with missing glucose measurements. The effectiveness of the proposed adaptive MPC based on rPLS is investigated with in silico subjects of the multivariable glucose-insulin-physiological variables simulator (mGIPsim). Results: The efficacy of the proposed adaptive MPC strategy in regulating the blood glucose concentration (BGC) of people with T1DM is assessed using the average percent time in range (TIR) for glucose, defined as 70 to 180 mg/dL inclusive, and the average percent time in hypoglycemia (54 mg/dL) and level 2 hypoglycemia (≤54 mg/dL). The TIR for a cohort of 20 virtual subjects of mGIPsim is 81.9% ± 7.4% (with no hypoglycemia or severe hypoglycemia) for the proposed MPC compared with 73.9% ± 7.6% (0.2% ± 0.1% in hypoglycemia and 0.1% ± 0.1% in level 2 hypoglycemia) for an MPC based on a recursive autoregressive exogenous (ARX) model. Conclusions: The adaptive MPC algorithm that incorporates prior knowledge in the recursive updating of the glucose prediction model can contribute to the development of fully automated artificial pancreas systems that can mitigate meal and physical activity disturbances.

Introduction: Diagnosis of T2DM necessitates clinical tests that are time-consuming and expensive. Machine learning (ML) techniques can accelerate the diagnosis and classification of T2DM and allow clinicians to personalize treatments based on blood glucose concentrations (BGC) , physical fitness (PF) , and diabetes distress patterns observed in daily life. Analyzing electronic health records (EHR) , physiological variables collected with wearable devices, and patient-reported outcomes (PROs) using ML techniques can lead to the development of clinical decision support tools that provide a comprehensive picture of an individual’s diabetes management needs. Methods: Clinical experimental data (n=85, F:40/M:45, HbA1c: 7.83±2.16; age: 57±7.72: BMI: 33.kg/m2±6.72; means and SDs) were used to identify clusters of subjects based on medical tests, and ML models were developed using readily measured data to classify new subjects to the identified clusters. Latent variable methods and k-means clustering were used to identify clusters based on HbA1c, physical performance tests, and PROs. ML models, including logistic regression (LR) and support vector machines (SVM) , were developed to assign new subjects to the identified clusters using the readily measured input variables from wearable devices, EHR and PROs. Results: Three distinct subject clusters were identified within the study cohort based on the descriptive variables. New subjects were assigned to the identified clusters with 87% and 91% accuracy for LR and SVM, respectively. Conclusion: EHRs, wearable device data, and PROs can be used to accurately and conveniently identify a person’s overall BGC, PF, and diabetes distress to aid in clinical decision-making. In the future, clinical decision support tools can be developed for personalized treatment suggestions based on cluster membership. Disclosure A.Shahidehpour: None. C.Fritschi: None. M.Rashid: None. A.Cinar: None. L.T.Quinn: n/a.

Objective: The 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 the detection of unannounced disturbances of 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 the 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 outperform 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.

A model predictive control (MPC) system based on latent variables (LV) model generated by using partial least squares (PLS) method is developed. The difference in the performance of MPCs that use recursively updated LV models based on autoregressive time series modeling (with exogenous inputs - ARX) and PLS is studied. The effect of signal noise on MPC performance is also investigated for both types of models. MPC performance is evaluated by regulating the blood glucose concentration (BGC) of people with Type 1 diabetes mellitus (T1DM) in simulation studies. Signal noise in glucose concentration sensor data, delays caused by insulin absorption and action, and disturbances caused by consumption of meals make the regulation of BGC difficult. The proposed controller is evaluated with 10 in-silico adult subjects of the UVa/Padova simulator with different levels of signal noise. The results illustrate the effectiveness of the MPC based on LV model. The average time for BGC in the safe range (70-180 mg/dL) for the LV-based MPC is 83.23% compared to 79.68% for the MPC based on ARX model when intravenous BGC values are used. The average time in safe range decreases to 76.04% and 71.92%, respectively, when using the generic CGM sensor of the simulator. It is reduced further to 71.93% and 67.20% when additional noise is added to CGM readings.

A model predictive control (MPC) formulation for a mammalian cell fed-batch bioreactor processes is developed. A nonlinear fundamental model for the bioreactor is used to generate a database of historical runs comprising of the measurement variables and the manipulated input feed flow rate to the bioreactor. The database is used with subspace identification methods to develop a state-space model of the process. The identified model is used to design various MPC formulations with different objective criteria, including the conventional trajectory-tracking objective function and a novel terminal objective for maximizing the product yield at completion of a run. Case studies involving the simulated bioreactor process demonstrate the efficacy of the MPC algorithms subject to unknown disturbances, random variations in the inlet feed glucose and glutamine concentrations, and measurement noise. Compared to the traditional proportional-integral control algorithm, the trajectory-tracking predictive control algorithm is able to better track the reference glucose concentration set-point with an improvement of 5.1% in the tracking error. The critical quality attribute predictive control algorithm designed to maximize the product yield results in a 3.9% increase in the product concentration at the completion of the run.

Wearable devices continuously measure multiple physiological variables to inform users of health and behavior indicators. The computed health indicators must rely on informative signals obtained by processing the raw physiological variables with powerful noise- and artifacts-filtering algorithms. In this study, we aimed to elucidate the effects of signal processing techniques on the accuracy of detecting and discriminating physical activity (PA) and acute psychological stress (APS) using physiological measurements (blood volume pulse, heart rate, skin temperature, galvanic skin response, and accelerometer) collected from a wristband. Data from 207 experiments involving 24 subjects were used to develop signal processing, feature extraction, and machine learning (ML) algorithms that can detect and discriminate PA and APS when they occur individually or concurrently, classify different types of PA and APS, and estimate energy expenditure (EE). Training data were used to generate feature variables from the physiological variables and develop ML models (na&iuml, ve Bayes, decision tree, k-nearest neighbor, linear discriminant, ensemble learning, and support vector machine). Results from an independent labeled testing data set demonstrate that PA was detected and classified with an accuracy of 99.3%, and APS was detected and classified with an accuracy of 92.7%, whereas the simultaneous occurrences of both PA and APS were detected and classified with an accuracy of 89.9% (relative to actual class labels), and EE was estimated with a low mean absolute error of 0.02 metabolic equivalent of task (MET).The data filtering and adaptive noise cancellation techniques used to mitigate the effects of noise and artifacts on the classification results increased the detection and discrimination accuracy by 0.7% and 3.0% for PA and APS, respectively, and by 18% for EE estimation. The results demonstrate the physiological measurements from wristband devices are susceptible to noise and artifacts, and elucidate the effects of signal processing and feature extraction on the accuracy of detection, classification, and estimation of PA and APS.

An adaptive-learning model predictive control (AL-MPC) framework is proposed for incorporating disturbance prediction, model uncertainty quantification, pattern learning, and recursive subspace identification for use in controlling complex dynamic systems with periodically recurring large random disturbances. The AL-MPC integrates online learning from historical data to predict the future evolution of the model output over a specified horizon and proactively mitigate significant disturbances. This goal is accomplished using dynamic regularized latent variable regression (DrLVR) approach to quantify disturbances from the past data and forecast their future progression time series. An enveloped path for the future behavior of the model output is extracted to further enhance the robustness of the closed-loop system. The controller set-point, penalty weights of the objective function, and constraints criteria can be modified in advance for the expected periods of the disturbance effects. The proposed AL-MPC is used to regulate glucose concentration in people with Type 1 diabetes by an automated insulin delivery system. Simulation results demonstrate the effectiveness of the proposed technique by improving the performance indices of the closed-loop system. The MPC algorithm integrated with DrLVR disturbance predictor has compared to MPC reinforced with dynamic principal component analysis linked with K-nearest neighbors and hyper-spherical clustering (k-means) technique. The simulation results illustrate that the AL-MPC can regulate the glucose concentrations of people with Type 1 diabetes to stay in the desired range (70–180) mg/dL 84.4% of the time without causing any hypoglycemia and hyperglycemia events.

This paper addresses the problem of heart rate (HR) monitoring from photo-plethysmography(PPG) sensors, where artifacts caused by body movements drastically affect the quality of the measurement signal. The PPG signal is windowed into consecutive segments, and for each time-windows, a Butterworth bandpass filter is utilized to attenuate high-frequency noises. Then, the PPG signal is processed by using the singular spectrum analysis technique to obtain a smooth PPG signal. In order to remove artifacts caused by the physical activity of the subject, the 3-dimensional accelerometer signal is used as an auxiliary signal to detect the presence of motion artifact (MA). A new spectral subtraction approach is proposed for MA rejection. For the purpose of HR estimation from the PPG signal, a feature extraction method is performed, and neural network binary classifier is used to detect the most probable frequencies corresponding to the actual HR. HR estimations are passed through a Kalman filter to result in smooth and accurate HR estimations.

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.

This paper presents the development of virtual patients to enable the simulation evaluation and assessment of multivariable control algorithms for biomedical systems. The virtual patients are generated by fitting the parameters of the models to clinical experimental data, followed by the estimation of the multivariate distribution of the actual patient parameters. The estimated multivariate distribution is then incorporated with constraints to ensure the sampling of synthetic virtual patients conforms to the actual patient parameter bounds. The sampled synthetic virtual patients are analyzed through multivariate statistical techniques and data clustering algorithms to prune out virtual subjects with similar characteristics or unrealistic dynamics, yielding a virtual patient population that is diverse and with individually distinct characteristics. The generated virtual patient population is used to evaluate multivariable nonlinear and adaptive control algorithms for insulin dosing in people with 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.

Insufficient vascularization limits the volume and complexity of biomaterials-based tissue engineering approaches. The formation of new blood vessels (neovascularization) is regulated by a complex interplay of cellular interactions with biochemical and biophysical signals provided by the extracellular matrix (ECM) which necessitates the development of biomaterial approaches that enable systematic modulation in matrix properties. Poly(ethylene) glycol-based hydrogel scaffolds were engineered with a range of decoupled and combined variations in integrin-binding peptide (RGD) ligand concentration, elastic modulus and proteolytic degradation rate using free-radical polymerization chemistry. The modularity of this system enabled a full factorial experimental design to simultaneously investigate the individual and interaction effects of these matrix cues on vascular sprout formation in 3D culture. Enhancements in scaffold proteolytic degradation rate promoted significant increases in vascular sprout length and junction number while increases in modulus significantly and negatively impacted vascular sprouting. We also observed that individual variations in immobilized RGD concentration did not significantly impact 3D vascular sprouting. Our findings revealed a previously unidentified and optimized combination whereby increases in both immobilized RGD concentration and proteolytic degradation rate resulted in significant and synergistic enhancements in 3D vascular spouting. The above-mentioned findings would have been challenging to uncover using one-factor-at-time experimental analyses.

Artifacts are disturbances affecting a measured signal that are not originating from the process itself. This paper addresses the problem of heart rate (HR) monitoring from a photoplethysmography (...

During portal vein embolization (PVE) prior to hepatectomy, the viability of the future liver remnant (FLR) is a critical determinant associated with the risk of post-operative liver failure and death.1 Care is taken to prevent embolization of viable liver and to prevent spread of embolization material. Alongside a review of the literature, we present a potentially devastating case of pulmonary embolization of material following PVE and liver resection.

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 withheld 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.

The glucose response to physical activity for a person with type 1 diabetes (T1D) depends upon the intensity and duration of the physical activity, plasma insulin concentrations, and the individual physical fitness level. To accurately model the glycemic response to physical activity, these factors must be considered.Several physiological models describing the glycemic response to physical activity are proposed by incorporating model terms proportional to the physical activity intensity and duration describing endogenous glucose production (EGP), glucose utilization, and glucose transfer from the plasma to tissues. Leveraging clinical data of T1D during physical activity, each model fit is assessed.The proposed model with terms accommodating EGP, glucose transfer, and insulin-independent glucose utilization allow for an improved simulation of physical activity glycemic responses with the greatest reduction in model error (mean absolute percentage error: 16.11 ± 4.82 vs. 19.49 ± 5.87, p = 0.002).The development of a physiologically plausible model with model terms representing each major contributor to glucose metabolism during physical activity can outperform traditional models with physical activity described through glucose utilization alone. This model accurately describes the relation of plasma insulin and physical activity intensity on glucose production and glucose utilization to generate the appropriately increasing, decreasing or stable glucose response for each physical activity condition. The proposed model will enable the in silico evaluation of automated insulin dosing algorithms designed to mitigate the effects of physical activity with the appropriate relationship between the reduction in basal insulin and the corresponding glycemic excursion.

Advisory and decision support systems can inundate users with frequent recommendations that are computationally expensive to generate. In this work, we develop a novel event-triggered insulin dosing algorithm for regulating glucose concentrations in people with Type 1 diabetes that is only executed when specified criteria are met. The proposed event-triggered insulin dosing algorithm is designed to function without announcements for meals and physical activity from the users. The algorithms are robust to missing data and signal dropouts that are common in free-living conditions. The safety and efficacy of the proposed event-triggered insulin dosing algorithm are demonstrated with simulation case studies that show the glucose concentrations are within the safe target range for 81.25% of the time, without any hypoglycemia, and the algorithm is only activated 62.67% of the time. The proposed computationally-efficient and robust event-triggered insulin dosing algorithm provides personalized assessment and can regulate glucose concentrations effectively in the presence of unannounced meals and physical activity.

Objective: Continuous glucose monitoring (CGM) enables prediction of the future glucose concentration (GC) trajectory for making informed diabetes management decisions. The glucose concentration values are affected by various physiological and metabolic variations, such as physical activity (PA) and acute psychological stress (APS), in addition to meals and insulin. In this work, we extend our adaptive glucose modeling framework to incorporate the effects of PA and APS on the GC predictions. Methods: A wristband conducive of use by free-living ambulatory people is used. The measured physiological variables are analyzed to generate new quantifiable input features for PA and APS. Machine learning techniques estimate the type and intensity of the PA and APS when they occur individually and concurrently. Variables quantifying the characteristics of both PA and APS are integrated as exogenous inputs in an adaptive system identification technique for enhancing the accuracy of GC predictions. Data from clinical experiments illustrate the improvement in GC prediction accuracy. Results: The average mean absolute error (MAE) of one-hour-ahead GC predictions with testing data decreases from 35.1 to 31.9 mg/dL (p-value = 0.01) with the inclusion of PA information, and it decreases from 16.9 to 14.2 mg/dL (p-value = 0.006) with the inclusion of PA and APS information. Conclusion: The first-ever glucose prediction model is developed that incorporates measures of physical activity and acute psychological stress to improve GC prediction accuracy. Significance: Modeling the effects of physical activity and acute psychological stress on glucose concentration values will improve diabetes management and enable informed meal, activity and insulin dosing decisions.

An adaptive and personalized multivariable artificial pancreas (mAP) system using plasma insulin estimates is proposed to efficiently accommodate major disturbances to the blood glucose concentration, such as meal and physical activity. Accurate adaptive glycemic models are developed through a recursive subspace identification technique with wearable physiological measurements and estimates of unannounced meal effect and plasma insulin concentration (PIC) along with continuous glucose concentration signals to characterize the glucose concentration dynamics under various conditions such as food consumption and physical activity. The identified models with time-varying parameters are employed in the design of an adaptive model predictive control (MPC) system that is cognizant of the PIC. The adaptive controller parameters, dynamic PIC constraint, addition of physiological measurements from wearable devices, feature variables generated from the glucose measurements, and estimation of uncertain model parameters, including the meal effect, enable the mAP system to effectively compute the optimal insulin infusion over diverse diurnal variations without meal and exercise announcements. Simulation case studies using a multivariable simulator demonstrate the efficacy of the proposed mAP system.

A controller-performance-assessment algorithm is developed to analyze the closed-loop behavior and modify the parameters of a control system employed in automated insulin delivery. To this end, various performance indices are defined to quantitatively evaluate the controller efficacy in real-time. The controller assessment and modification module also incorporates online learning from historical data to anticipate impending disturbances and proactively counteract their effects. A dynamic safety constraint derived from estimates of the physiological states ensures safety of the controlled drug dosing. Using a multivariable simulation platform for type 1 diabetes mellitus, the controller assessment and modification module is applied to the problem of regulating glucose concentrations in people with diabetes by means of automated insulin delivery with an artificial pancreas, and the results demonstrate the improvement in controller performance using the performance-assessment module.

Background: Physical activity presents a significant challenge for glycemic control in individuals with type 1 diabetes. As accurate glycemic predictions are key to successful automated decision-making systems (eg, artificial pancreas, AP), the inclusion of additional physiological variables in the estimation of the metabolic state may improve the glucose prediction accuracy during exercise. Methods: Predictor-based subspace identification is applied to a dynamic glucose prediction model including heart rate measurements along with variables representing the carbohydrate consumption and insulin boluses. To demonstrate the improvement in prediction ability due to the additional heart rate variable, the performance of the proposed modeling technique is evaluated with (SID-HR) and without heart rate (SID-2) as an additional input using experimental data involving adolescents at ski camp. Furthermore, the performance of the proposed approach is compared to that of the metabolic state observer (MSO) model currently used in the University of Virginia AP algorithm. Results: The addition of heart rate in the subspace-based model (SID-HR) yields a statistically significant improvement in the root-mean-square error compared to the SID-2 model ( P < .001) and the standard MSO ( P < .001). Furthermore, the SID-HR model performed favorably in comparison to the SID-2 and MSO models after accounting for its increased complexity. Conclusions: Directly considering the effects of physical activity levels on glycemic dynamics through the inclusion of heart rate as an additional input variable in the glucose dynamics model improves the glucose prediction accuracy. The proposed methodology could improve exercise-informed model-based predictive control algorithms in artificial pancreas systems.

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.

Automated drug delivery can reduce many repetitive manual tasks and human error in the treatment of chronic diseases. The automation can range from duplication of the procedures conducted by the human user to leveraging machine learning (ML) and advanced control techniques to provide a comprehensive automated system that can personalize the treatment decisions, take into account the habits and predicted future activities of the user, and leverage measurable disturbances to enhance the accuracy of control decisions. This potential improvement can only be achieved by developing a set of algorithms that can handle the nonlinearities and time-varying nature of the relevant metabolic processes. Regulation of blood glucose concentration provides an excellent case for illustrating the role of various ML, estimation and automatic control techniques to achieve good glucose control in spite of disturbances such as meals, physical activities, and physiological stress. The chapter will outline various control philosophies and estimators that can provide valuable information about variables that cannot be measured in real time such as plasma insulin concentration (PIC) and the potential impact of physical activities and acute psychological stress. PIC estimates can be used to adjust the aggressiveness of the controller, while physical activity and stress estimates enable feedforward control to compensate for impending changes in glucose levels before these changes manifest themselves in glucose readings. Time-varying glycemic models obtained through a recursive system identification by using continuous glucose measurements, physiological measurements, estimates of unannounced meal effects, and PIC form the foundation for adaptive control. An adaptive multivariable model predictive control algorithm illustrates the use of fully automated control that does not require any manual inputs from users, reducing the treatment burden of the chronic disease.

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.

Developing viable and robust software is an inescapable artifact of graduate research. The challenges lie in the complexities of developing, deploying, and securing software to support the research objectives. Combined with the transitive nature of students, the management of the software development and launch process is an arduous task. A standardized framework for developing and launching complex software is required. Within a university, individual departments do not typically possess the expertise, resources, software and infrastructure to translate research results to a viable product or tool. Extending upon the research of Hilton et al., [5] we designed a software development pipeline in an integrated multi-disciplinary research context. The integrated and collaborative software pipeline formulated from the onset of the project streamlines the development phase and provides an iterative feedback and testing environment. This approach is applied to the development of automated insulin delivery systems, with the synergistic efforts of interdisciplinary teams yielding a mobile application and server software solutions, and a framework for the iterative advancement of the software capabilities into the future.

Background: Multivariable artificial pancreas (mAP) systems are developed to supplement the continuous glucose monitoring (CGM) data with physiological measurements from wearable devices and automatically incorporate physical activity information in proactive insulin dose decisions. The wearable devices incorporated into mAP systems collect data in real-time to determine the type and intensity of the physical activity without requiring manual entries. As the first mAP, our multivariable Automated Insulin Delivery (mAID) system does not require user input for meals or exercise. In silico studies and clinical experiments have shown improvement in glycemic control with mAID during unannounced physical activity compared to the glucose-only AP systems. Methods: mAID is evaluated in three clinical experiments, with one experiment conducted in a crossover fashion where the glucose-only AP is compared with the mAID. The experiments involved two subjects (1 male and 1 female) with T1DM, age 35 - 45 years, bodyweight 70 - 110 kg, total daily insulin of 35 - 80 U, and insulin-to-carb ratio 6.5 - 7 g/U. The experiments included meals and medium-intensity exercises (treadmill run and stationary bike) without any announcements to the system. The target glycemic range was [70, 180] mg/dL. Results: The mAID system regulated glycemia more effectively, than the glucose-only AP, which had 13.33% of time spend below 70 mg/dL and 86.67% of time spent in the target range, with a significant 55 g of rescue carbs given to avert impending low glucose. The mAID system resulted in all time spent in the target range for the same subject. Across the closed-loop experiments, subjects spent a total of 87.69% of time in the target range. Conclusions: The mAID system supplements the CGM data with real-time physiological data to automatically handle physical activity and maintain tight glycemic control. Disclosure M. Rashid: None. M. Sevil: None. N. Hobbs: None. I. Hajizadeh: None. M. Askari: None. R. Brandt: None. M. Park: None. L.T. Quinn: None. A. Cinar: Research Support; Self; Dexcom, Inc., JDRF. Stock/Shareholder; Self; Dexcom, Inc., Eli Lilly and Company, Insulet Corporation, Medtronic, Novo Nordisk A/S, Tandem Diabetes Care. Funding National Institutes of Health (1DP3DK101075-01, 1DP3DK101077-01); JDRF (A18-0036-001)

A personalized multivariable, multimodel artificial pancreas (PMM-AP) system is developed to automate and personalize insulin treatment of type 1 diabetes. The proposed PMM-AP is a fully automated insulin delivery system that works with no meal and physical activity announcements. An adaptive-personalized plasma insulin concentration (PIC) estimator is designed to quantify the amount of active insulin present in the body. Time-varying glycemic models are obtained through a recursive system identification technique using physiological measurements, continuous glucose measurements, estimates of unannounced meal effects and PIC. An adaptive-learning model predictive control algorithm is then designed using the identified glycemic models and PIC estimates to compute a safe and optimal insulin amount. Simulation case studies demonstrate the performance of the proposed PMM-AP system.

Closed-loop control of glucose concentrations in type 1 diabetes mellitus has progressed significantly over the last decade. Automated insulin delivery systems, or artificial pancreas systems, have advanced to incorporate model-based predictive controllers developed on the basis of adaptive and personalized glucose-insulin models. This chapter reviews the advancements and the state of the art in control algorithms and modeling paradigms employed to develop the artificial pancreas systems. Recent technological breakthroughs in dual-hormone systems, run-to-run control approaches, and multivariable control architectures for improving glycemic control are discussed. A novel integrated physiological and metabolic simulation platform for multivariable glucose, insulin, and physiological variables simulation is used to demonstrate the ability of the artificial pancreas systems to counteract the diverse meal and exercise disturbances and maintain euglycemia.

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.

Linear empirical dynamic models have been widely used for blood glucose prediction and risks prevention in people with type 1 diabetes. More accurate blood glucose prediction models with longer prediction horizon (PH) are desirable to enable warnings to patients about imminent blood glucose changes with enough time to take corrective actions. In this study, a blood glucose prediction method is developed by integrating the predictions of a set of seasonal local models (each of them corresponding to different glucose profiles observed along historical data). In the modeling step, the number of sets and their corresponding glucose profiles characteristics are obtained by clustering techniques (Fuzzy C-Means). Then, Box-Jenkins methodology is used to identify a seasonal model for each set. Finally, blood glucose predictions of local models are integrated using different techniques. The proposed method is tested by using 18 60-h closed-loop experiments (including different exercise types and artificial pancreas strategies) and achieving mean absolute percentage error (MAPE) of 2.94%, 3.89%, 5.41%, 6.29% and 8.66% for 15-, 30-, 45-, 60-, and 90-min PHs, respectively.

Background: Despite the recent advancements in the modeling of glycemic dynamics for type 1 diabetes mellitus, automatically considering unannounced meals and exercise without manual user inputs remains challenging. Method: An adaptive model identification technique that incorporates exercise information and estimates of the effects of unannounced meals obtained automatically without user input is proposed in this work. The effects of the unknown consumed carbohydrates are estimated using an individualized unscented Kalman filtering algorithm employing an augmented glucose-insulin dynamic model, and exercise information is acquired from noninvasive physiological measurements. The additional information on meals and exercise is incorporated with personalized estimates of plasma insulin concentration and glucose measurement data in an adaptive model identification algorithm. Results: The efficacy of the proposed personalized and adaptive modeling algorithm is demonstrated using clinical data involving closed-loop experiments of the artificial pancreas system, and the results demonstrate accurate glycemic modeling with the average root-mean-square error (mean absolute error) of 25.50 mg/dL (18.18 mg/dL) for six-step (30 minutes ahead) predictions. Conclusions: The approach presented is able to identify reliable time-varying individualized glucose-insulin models.

Background: The artificial pancreas (AP) system, a technology that automatically administers exogenous insulin in people with type 1 diabetes mellitus (T1DM) to regulate their blood glucose concentrations, necessitates the estimation of the amount of active insulin already present in the body to avoid overdosing. Method: An adaptive and personalized plasma insulin concentration (PIC) estimator is designed in this work to accurately quantify the insulin present in the bloodstream. The proposed PIC estimation approach incorporates Hovorka’s glucose-insulin model with the unscented Kalman filtering algorithm. Methods for the personalized initialization of the time-varying model parameters to individual patients for improved estimator convergence are developed. Data from 20 three-days-long closed-loop clinical experiments conducted involving subjects with T1DM are used to evaluate the proposed PIC estimation approach. Results: The proposed methods are applied to the clinical data containing significant disturbances, such as unannounced meals and exercise, and the results demonstrate the accurate real-time estimation of the PIC with the root mean square error of 7.15 and 9.25 mU/L for the optimization-based fitted parameters and partial least squares regression-based testing parameters, respectively. Conclusions: The accurate real-time estimation of PIC will benefit the AP systems by preventing overdelivery of insulin when significant insulin is present in the bloodstream.

Background: Automatically attenuating the postprandial rise in the blood glucose concentration without manual meal announcement is a significant challenge for artificial pancreas (AP) systems. In this study, a meal module is proposed to detect the consumption of a meal and to estimate the amount of carbohydrate (CHO) intake. Methods: The meals are detected based on qualitative variables describing variation of continuous glucose monitoring (CGM) readings. The CHO content of the meals/snacks is estimated by a fuzzy system using CGM and subcutaneous insulin delivery data. The meal bolus amount is computed according to the patient's insulin to CHO ratio. Integration of the meal module into a multivariable AP system allows revision of estimated CHO based on knowledge about physical activity, sleep, and the risk of hypoglycemia before the final decision for a meal bolus is made. Results: The algorithm is evaluated by using 117 meals/snacks in retrospective data from 11 subjects with type 1 diabetes. Sensitivity, defined as the percentage of correctly detected meals and snacks, is 93.5% for meals and 68.0% for snacks. The percentage of false positives, defined as the proportion of false detections relative to the total number of detected meals and snacks, is 20.8%. Conclusions: Integration of a meal detection module in an AP system is a further step toward an automated AP without manual entries. Detection of a consumed meal/snack and infusion of insulin boluses using an estimate of CHO enables the AP system to automatically prevent postprandial hyperglycemia.