The UVA/Padova Type 1 Diabetes Simulator Goes From Single Meal to Single Day

Time to Moderate and Severe Hyperglycemia and Ketonemia Following an Insulin Pump Occlusion

INTRODUCTION

Insulin pump therapy can be adversely affected by interruption of insulin flow, leading to a rise in blood glucose (BG) and subsequently of blood beta-hydroxybutyrate (BHB) ketone levels.

METHODS

We performed a PubMed search for English language reports (January 1982 to July 2024) estimating the rate of rise in BG and/or BHB after ≥ 60 minutes of interruption of continuous subcutaneous insulin infusion (CSII) in persons with type 1 diabetes (PwT1D). We also simulated the rise in BG in a virtual population of 100 adults with T1D following suspension of continuous subcutaneous insulin infusion.

RESULTS

We identified eight relevant studies where BG and BHB (seven of these eight studies) were measured following suspension of CSII as a model for occlusion. After 60 minutes post-suspension, the mean extracted rates of rise averaged 0.62 mg/dL/min (37 mg/dL/h) for BG and 0.0038 mmol/L/min (0.20 mmol/L/h) for BHB. Mean estimated time to moderately/severely elevated BG (300/400 mg/dL) or BHB (1.6/3.0 mmol/L) was, respectively, 5.8/8.5 and 8.0/14.2 hours. The simulation model predicted moderately/severely elevated BG (300/400 mg/dL) after 9.25/12, 6.75/8.75, and 4.75/5.75 hours in the virtual subjects post-interruption with small (5th percentile), medium (50th percentile), and large (95th percentile) hyperglycemic changes.

DISCUSSION

Clinical studies and a simulation model similarly predicted that, following CSII interruption, moderate/severe hyperglycemia can occur within 5-9/6-14 hours, and clinical studies predicted that moderate/severe ketonemia can occur within 7-12/13-21 hours. Patients and clinicians should be aware of this timing when considering the risks of developing metabolic complications after insulin pump occlusion.

A GENERATIVE APPROACH TO TESTING THE PERFORMANCE OF PHYSIOLOGICAL CONTROL ALGORITHMS

Background

Physiological closed-loop control algorithms play an important role in the development of autonomous medical care systems, a promising area of research that has the potential to deliver healthcare therapies meeting each patient's specific needs. Computational approaches can support the evaluation of physiological closed-loop control algorithms considering various sources of patient variability that they may be presented with.

Method of Approach

In this paper, we present a generative approach to testing the performance of physiological closed-loop control algorithms. This approach exploits a generative physiological model (which consists of stochastic and dynamic components that represent diverse physiological behaviors across a patient population) to generate a select group of virtual subjects. By testing a physiological closed-loop control algorithm against this select group, the approach estimates the distribution of relevant performance metrics in the represented population. We illustrate the promise of this approach by applying it to a practical case study on testing a closed-loop fluid resuscitation control algorithm designed for hemodynamic management.

Results

In this context, we show that the proposed approach can test the algorithm against virtual subjects equipped with a wide range of plausible physiological characteristics and behavior, and that the test results can be used to estimate the distribution of relevant performance metrics in the represented population.

Conclusions

In sum, the generative testing approach may offer a practical, efficient solution for conducting pre-clinical tests on physiological closed-loop control algorithms.

Auto adaptation of closed-loop insulin delivery system using continuous reward functions and incremental discretization

Several closed or hybrid loop controllers for Blood Glucose (BG) regulation, which are also known as Artificial Pancreas (AP) Systems or Automated Insulin Delivery systems (AIDs), are in development worldwide. Most AIDs are designed and evaluated for short-term performance, with a particular emphasis on the post-meal period. However, if controllers are not adapted properly to account for variations in physiology that affect Insulin Sensitivity (IS), the AIDs may perform inadequately. In this work, the performance of two Reinforcement Learning (RL) agents trained under both piecewise and continuous reward functions is evaluated in-silico for long-term adaptation of a Fully Automated Insulin Delivery (fAID) system. An automatic adaptive discretization scheme that expands the state space as needed is also implemented to avoid disproportionate state space exploration. The proposed agents are evaluated for long-term adaptation of the Automatic Regulation of Glucose (ARG) algorithm, considering variations in IS. Results show that both RL agents have improved performance compared to a rule-based decision-making approach and the baseline controller for the majority of the adult population. Moreover, the use of a continuous shaped reward function proves to enhance the performance of the agents further than a piecewise one.

Digital Twins in Type 1 Diabetes: A Systematic Review

Digital twin is a new concept that is rapidly gaining recognition especially in the medical field. Indeed, being a virtual representation of real-world entities and processes, a digital twin can be used to accurately represent the patients' disease, clarify the treatment target, and realize personalized and precise therapies. However, despite being a revolutionary concept, the diffusion of digital twins in type 1 diabetes (T1D) is still limited. In this systematic review, we analyzed structure, operating conditions, and characteristics of digital twins being developed for T1D. Our search covered published documents until March 2024: 220 publications were identified, 37 of which were duplicated entries; in addition, 173 publications were removed after inspection of titles, abstracts, and keywords; and finally, 11 publications were fully reviewed, of which 8 were deemed eligible for inclusion. We found that all eight methodologies are not comprehensive multi-scale virtual replicas of the individual with T1D, but they all focus on describing glucose-insulin metabolism, aiming to simulate glucose concentration resultant from therapeutic interventions. In this review, we will compare and analyze different factors characterizing these digital twins, such as operating principles (mathematical model, twinning procedure, validation and assessment) and the key aspects for practical adoption (inclusion of physical activity, data required for twinning, open-source availability). We will conclude the paper listing which, in our opinion, are the current limitations and future directives of digital twins in T1D, hoping that this article can be helpful to researchers working on diabetes technologies to further develop the use of such an important instrument.

Tailoring the Padova Type 2 Diabetes Simulator for Treatment Guidance in Target Populations

OBJECTIVE

The Padova type 2 diabetes (T2D) simulator (T2DS) has been recently proposed to optimize T2D treatments including novel long-acting insulins. It consists of a physiological model and an in silico population describing glucose dynamics, derived from early-stage T2D subjects studied with sophisticated tracer-based experimental techniques. This limits T2DS domain of validity to this specific sub-population. Conversely, running simulations in insulin-naïve or advanced T2D subjects, would be more valuable. However, it is rarely possible or cost-effective to run complex experiments in such populations. Therefore, we propose a method for tuning the T2DS to any desired T2D sub-population using published clinical data. As case study, we extended the T2DS to insulin-naïve T2D subjects, who need to start insulin therapy to compensate the reduced insulin function.

METHODS

T2DS model was identified based on literature data of the target population. The estimated parameters were used to generate a virtual cohort of insulin-naïve T2D subjects (inC1). A model of basal insulin degludec (IDeg) was also incorporated into the T2DS to enable basal insulin therapy. The resulting tailored T2DS was assessed by simulating IDeg therapy initiation and comparing simulated vs. clinical trial outcomes. For further validation, this procedure was reiterated to generate a new cohort of insulin-naïve T2D (inC2) assuming inC1 as target population.

RESULTS

No statistically significant differences were found when comparing fasting plasma glucose and IDeg dose, neither in clinical data vs. inC1, nor inC1 vs. inC2.

CONCLUSIONS

The tuned T2DS allowed reproducing the main findings of clinical studies in insulin-naïve T2D subjects.

SIGNIFICANCE

The proposed methodology makes the Padova T2DS usable for supporting treatment guidance in target T2D populations.

A strategy based on integer programming for optimal dosing and timing of preventive hypoglycemic treatments in type 1 diabetes management

BACKGROUND AND OBJECTIVES

One of the major problems related to type 1 diabetes (T1D) management is hypoglycemia, a condition characterized by low blood glucose levels and responsible for reduced quality of life and increased mortality. Fast-acting carbohydrates, also known as hypoglycemic treatments (HT), can counteract this event. In the literature, dosage and timing of HT are usually based on heuristic rules. In the present work, we propose an algorithm for mitigating hypoglycemia by suggesting preventive HT consumption, with dosages and timing determined by solving an optimization problem.

METHODS

By leveraging integer programming and linear inequality constraints, the algorithm can bind the amount of suggested carbohydrates to standardized quantities (i.e., those available in "off-the-shelf" HT) and the minimal distance between consecutive suggestions (to reduce the nuisance for patients).

RESULTS

The proposed method was tested in silico and compared with competitor algorithms using the UVa/Padova T1D simulator. At the cost of a slight increase of HT consumed per day, the proposed algorithm produces the lowest median and interquartile range of the time spent in hypoglycemia, with a statistically significant improvement over most competitor algorithms. Also, the average number of hypoglycemic events per day is reduced to 0 in median.

CONCLUSIONS

Thanks to its positive performances and reduced computational burden, the proposed algorithm could be a candidate tool for integration in a DSS aimed at improving T1D management.

JL, Garcia-Tirado J, DeBoer MD, Breton MD. Adjusting Therapy Profiles When Switching to Ultra-Rapid Lispro in an Advanced Hybrid Closed-Loop System: An in Silico Study

BACKGROUND

It has been shown that insulin acceleration by itself might not be sufficient to see clear improvements in glycemic metrics, and insulin therapy may need to be adjusted to fully leverage the extra safety margin provided by faster pharmacokinetic (PK) and pharmacodynamic (PD) profiles. The objective of this work is to explore how to perform such adjustments on a commercially available automated insulin delivery (AID) system.

METHODS

Ultra-rapid lispro (URLi) is modeled within the UVA/Padova simulation platform using data from previously published clamp studies. The Control-IQ AID algorithm is selected as it leverages carbohydrate-to-insulin ratio (CR in g/U), correction factor (CF in mg/dL/U), and basal rate (BR in U/h) daily profiles that are fully customizable. An experiment roadmap is proposed to understand how to safely modify these profiles when switching from lispro to URLi.

RESULTS

Simulations show that a 7% decrease in CR (approximately an 8% increase in prandial insulin) and a 7.5% increase in BR lead to cumulative improvements in glucose control with URLi. Comparing with baseline metrics using lispro, a clinically significant increase in time in the range of 70 to 180 mg/dL (overall: 70.2%-75.2%, P < .001; 6 am-12 am: 62.4%-68.5%, P < .001) and a reduction in time below 70 mg/dL (overall: 1.8%-1.2%, P < .001; 6 am-12 am: 1.8%-1.3%, P < .001) were observed.

CONCLUSION

Properly adjusting therapy parameters allows to fully leverage glucose control benefits provided by faster insulin analogues, opening opportunities to take another step forward into a next generation of more effective AID solutions.

Validation of the UVA Simulation Replay Methodology Using Clinical Data: Reproducing a Randomized Clinical Trial

Background: Computer simulators of human metabolism are powerful tools to design and validate new diabetes treatments. However, these platforms are often limited in the diversity of behaviors and glycemic conditions they can reproduce. Replay methodologies leverage field-collected data to create ad hoc simulation environments representative of real-life conditions. After formal validations of our method in prior publications, we demonstrate its capacity to reproduce a recent clinical trial. Methods: Using the replay methodology, an ensemble of replay simulators was generated using data from a randomized crossover clinical trial comparing the hybrid closed loop (HCL) and fully closed loop (FCL) control modalities in automated insulin delivery (AID), creating 64 subject/modality pairs. Each virtual subject was exposed to the alternate AID modality to compare the simulated versus observed glycemic outcomes. Equivalence tests were performed for time in, below, and above range (TIR, TBR, and TAR) and high and low blood glucose indices (HBGI and LBGI) considering equivalence margins corresponding to clinical significance. Results: TIR, TAR, LBGI, and HBGI showed statistical and clinical equivalence between the original and the simulated data; TBR failed the equivalence test. For example, in the HCL mode, simulated TIR was 84.89% versus an observed 84.31% (P = 0.0170, confidence interval [CI] [-3.96, 2.79]), and for FCL mode, TIR was 76.58% versus 77.41% (P = 0.0222, CI [-2.54, 4.20]). Conclusion: Clinical trial data confirm the prior in silico validation of the UVA replay method in predicting the glycemic impact of modified insulin treatments. This in vivo demonstration justifies the application of the replay method to the personalization and adaptation of treatment strategies in people with type 1 diabetes.

Supervised and Unsupervised Approaches for the Real-Time Detection of Undesired Insulin Suspension Caused by Malfunctions

BACKGROUND

Automated insulin delivery (AID) systems, permit improved treatment of type 1 diabetes (T1D). Unfortunately, malfunctioning in the insulin pump or in the infusion set can prevent insulin from being administered, reducing the AID efficacy and posing the patient at risk. Different data-driven methods available in the literature can be used to deal with the problem of automatically detecting complete insulin suspension in real-time. This article investigates both supervised and unsupervised strategies and proposes a fair comparison under either population or personalized settings.

METHODS

Several algorithms are compared using data generated through the UVA/Padova T1D simulator, a computer simulator widely used to test control strategies in silico and accepted by the Food and Drugs Administration (FDA) as a substitute to animal pre-clinical trials. Two synthetic data sets, each consisting of 100 virtual subjects monitored for 1 month, were generated. Occasional faults of the insulin pump are simulated as complete occlusions by suspending the therapy administration. Personalized algorithms are investigated with unsupervised approaches only, since personalized labels are hardly available.

RESULTS

In the population scenario, the supervised approach outperforms the unsupervised strategy. In particular, logistic regression and random forest achieves a recall of 72% and 82%, with 0.12 and 0.21 false positives (FP) per day, respectively. In the personalized setting scenario, the unsupervised algorithms are tailored on each patient and outperform the population ones, in particular isolation forest achieves a recall 80% and 0.06 FPs per day.

CONCLUSIONS

This article suggests that unsupervised personalized approach, by addressing the large variability in glucose response among individuals with T1D, is superior to other one-fits-all approaches in detecting insulin suspensions caused by malfunctioning. Population methodologies can be effectively used while waiting to collect sufficient patient data, when the system is installed on a new patient.

Integration of a Safety Module to Prevent Rebound Hypoglycemia in Closed-Loop Artificial Pancreas Systems

BACKGROUND

With automated insulin delivery (AID) systems becoming widely adopted in the management of type 1 diabetes, we have seen an increase in occurrences of rebound hypoglycemia or generated hypoglycemia induced by the controller's response to rapid glucose rises following rescue carbohydrates. Furthermore, as AID systems aim to enable complete automation of prandial control, algorithms are designed to react even more strongly to glycemic rises. This work introduces a rebound hypoglycemia prevention layer (HypoSafe) that can be easily integrated into any AID system.

METHODS

HypoSafe constrains the maximum permissible insulin delivery dose based on the minimum glucose reading from the previous hour and the current glucose level. To demonstrate its efficacy, we integrated HypoSafe into the latest University of Virginia (UVA) AID system and simulated two scenarios using the 100-adult cohort of the UVA/Padova T1D simulator: a nominal case including three unannounced meals, and another case where hypoglycemia was purposely induced by an overestimated manual bolus.

RESULTS

In both simulation scenarios, rebound hypoglycemia events were reduced with HypoSafe (nominal: from 39 to 0, hypo-induced: from 55 to 7) by attenuating the commanded basal (nominal: 0.27U vs. 0.04U, hypo-induced: 0.27U vs. 0.03U) and bolus (nominal: 1.02U vs. 0.05U, hypo-induced: 0.43U vs. 0.02U) within the 30-minute interval after treating a hypoglycemia event. No clinically significant changes resulted for time in the range of 70 to 180 mg/dL or above 180 mg/dL.

CONCLUSION

HypoSafe was shown to be effective in reducing rebound hypoglycemia events without affecting achieved time in range when combined with an advanced AID system.

Short- and Long-Term Effects on Glucose Control of Nonadherence to Insulin Therapy in People With Type 2 Diabetes An In Silico Study

BACKGROUND

Strict adherence to multiple daily insulin (MDI) therapy is a cornerstone for the achievement of good glucose control in people with advanced type 2 diabetes (T2D). Here, we aim to in silico assess glucose control in T2D subjects with poor adherence to MDI therapy.

METHODS

We tuned the Padova T2D Simulator, originally describing early-stage T2D physiology, around advanced T2D people. One hundred in silico advanced T2D subjects were generated and equipped with optimal MDI therapy: specifically, basal and bolus insulin amounts and injection times were individualized for each subject by applying titration algorithms that iteratively update insulin dose based on glucose deviation from its target. Then, the effect of nonadhering to MDI therapy was assessed using standard glucose control metrics calculated in two 6-month 3-meal/day in silico scenarios: in Scenario 1, subjects received the optimal basal and prandial insulin bolus at each meal; in Scenario 2, subjects received optimal basal insulin and randomly delayed or skipped the prandial insulin bolus in 3 lunches during working days and 1 dinner during weekends.

RESULTS

A statistically significant degradation was found in all glucose control outcome metrics in Scenario 2 versus Scenario 1: e.g., percent time above 180 mg/dL increased by 22.2% and glucose management index by 0.2%.

CONCLUSIONS

Impaired adherence to MDI therapy in T2D leads to glucose control deteriorations in both short and long terms. Interestingly, short-term hyperglycemia seems being contrasted by residual endogenous insulin secretion, which statistically increased by 3-fold after delayed/skipped insulin boluses compared with optimal ones.

Risk of hypoglycemia in type 1 diabetes management: An in-silico sensitivity analysis to assess and rank the quantitative impact of different behavioral factors

BACKGROUND AND OBJECTIVE

In type 1 diabetes (T1D), a quantitative evaluation of the impact on hypoglycemia of suboptimal therapeutic decision (e.g. incorrect estimation of the ingested carbohydrates, inaccurate insulin timing, etc) is unavailable. Clinical trials to measure sensitivity to patient actions would be expensive, exposed to confounding factors and risky for the participants. In this work, a T1D patient decision simulator (T1D-PDS), realistically reproducing blood glucose dynamics in a large virtual population, is used to perform extensive in-silico trials and the so-derived data employed to implement a sensitivity analysis that ranks different behavioral factors based on their impact on a clinically meaningful parameter, the time below range (TBR).

METHODS

Eleven behavioral factors impacting on hypoglycemia are considered. The T1D-PDS was used to perform multiple 2-week simulations involving 100 adults, by testing about 3500 different perturbations for nominal behavior. A local linear approximation of the function linking the TBR and the factors were computed to derive sensitivity indices (SIs), quantifying the impact of each factor on TBR variations.

RESULTS

The obtained ranking quantifies importance of factors w.r.t. the others. Factors apparently related to hypoglycemia were correctly placed on the top of the ranking, including systematic (SI=2.05%) and random (SI=1.35%) carb-counting error, hypotreatment dose (SI=-1.21%), insulin bolus time w.r.t. mealtime (SI=1.09%).

CONCLUSIONS

The obtained SIs allowed to rank behavioral factors based on their impact on TBR. The behavioral factors identified as most influential can be prioritized in patient training.

Blood glucose forecasting from temporal and static information in children with T1D

Background

The overarching goal of blood glucose forecasting is to assist individuals with type 1 diabetes (T1D) in avoiding hyper- or hypoglycemic conditions. While deep learning approaches have shown promising results for blood glucose forecasting in adults with T1D, it is not known if these results generalize to children. Possible reasons are physical activity (PA), which is often unplanned in children, as well as age and development of a child, which both have an effect on the blood glucose level.

Materials and Methods

In this study, we collected time series measurements of glucose levels, carbohydrate intake, insulin-dosing and physical activity from children with T1D for one week in an ethics approved prospective observational study, which included daily physical activities. We investigate the performance of state-of-the-art deep learning methods for adult data-(dilated) recurrent neural networks and a transformer-on our dataset for short-term (30  min) and long-term (2  h) prediction. We propose to integrate static patient characteristics, such as age, gender, BMI, and percentage of basal insulin, to account for the heterogeneity of our study group.

Results

Integrating static patient characteristics (SPC) proves beneficial, especially for short-term prediction. LSTMs and GRUs with SPC perform best for a prediction horizon of 30  min (RMSE of 1.66  mmol/l), a vanilla RNN with SPC performs best across different prediction horizons, while the performance significantly decays for long-term prediction. For prediction during the night, the best method improves to an RMSE of 1.50  mmol/l. Overall, the results for our baselines and RNN models indicate that blood glucose forecasting for children conducting regular physical activity is more challenging than for previously studied adult data.

Conclusion

We find that integrating static data improves the performance of deep-learning architectures for blood glucose forecasting of children with T1D and achieves promising results for short-term prediction. Despite these improvements, additional clinical studies are warranted to extend forecasting to longer-term prediction horizons.

Simulation-driven optimization of insulin therapy profiles in a commercial hybrid closed-loop system

BACKGROUND

Automated insulin delivery (AID) has represented a breakthrough in managing type 1 diabetes (T1D), showing safe and effective glucose control extensively across the board. However, metabolic variability still poses a challenge to commercial hybrid closed-loop (HCL) solutions, whose performance depends on customizable insulin therapy profiles. In this work, we propose an Identification-Replay-Optimization (IRO) approach to optimize gradually and safely such profiles for the Control-IQ AID algorithm.

METHODS

Closed-loop data are generated using the full adult cohort of the UVA/Padova T1D simulation platform in diverse glycemic scenarios. For each subject, daily records are processed and used to estimate a personalized model of the underlying insulin-glucose dynamics. Every two weeks, all identified models are integrated into an optimization procedure where daily basal and bolus profiles are adjusted so as to minimize the risks for hypo- and hyperglycemia. The proposed strategy is tested under different scenarios of metabolic and behavioral variability in order to evaluate the efficacy and convergence of the proposed strategy. Finally, glycemic metrics between cycles are compared using paired t-tests with p<0.05 as the significance threshold.

RESULTS

Simulations reveal that the proposed IRO approach was able to improve glucose control over time by safely mitigating the risks for both hypo- and hyperglycemia. Furthermore, smaller changes were recommended at each cycle, indicating convergence when simulation conditions were maintained.

CONCLUSIONS

The use of reliable simulation-driven tools capable of accurately reproducing field-collected data and predicting changes can substantially shorten the process of optimizing insulin therapy, adjusting it to metabolic changes and leading to improved glucose control.

Individualized Models for Glucose Prediction in Type 1 Diabetes: Comparing Black-Box Approaches to a Physiological White-Box One

OBJECTIVE

Accurate blood glucose (BG) prediction are key in next-generation tools for type 1 diabetes (T1D) management, such as improved decision support systems and advanced closed-loop control. Glucose prediction algorithms commonly rely on black-box models. Large physiological models, successfully adopted for simulation, were little explored for glucose prediction, mostly because their parameters are hard to individualize. In this work, we develop a BG prediction algorithm based on a personalized physiological model inspired by the UVA/Padova T1D Simulator. Then we compare white-box and advanced black-box personalized prediction techniques.

METHODS

A personalized nonlinear physiological model is identified from patient data through a Bayesian approach based on Markov Chain Monte Carlo technique. The individualized model was integrated within a particle filter (PF) to predict future BG concentrations. The black-box methodologies considered are non-parametric models estimated via gaussian regression (NP), three deep learning methods: long-short-term-memory (LSTM), gated recurrent unit (GRU), temporal convolutional networks (TCN), and a recursive autoregressive with exogenous input model (rARX). BG forecasting performances are assessed for several prediction horizons (PH) on 12 individuals with T1D, monitored in free-living conditions under open-loop therapy for 10 weeks.

RESULTS

NP models provide the most effective BG predictions by achieving a root mean square error (RMSE), RMSE = 18.99 mg/dL, RMSE = 25.72 mg/dL and RMSE = 31.60 mg/dL, significantly outperforming: LSTM, GRU (for PH = 30 minutes), TCN, rARX, and the proposed physiological model for PH=30, 45 and 60 minutes.

CONCLUSIONS

Black-box strategies remain preferable for glucose prediction even when compared to a white-box model with sound physiological structure and individualized parameters.

Developing the UVA/Padova Type 1 Diabetes Simulator: Modeling, Validation, Refinements, and Utility

Arguably, diabetes mellitus is one of the best quantified human conditions. In the past 50 years, the metabolic monitoring technologies progressed from occasional assessment of average glycemia via HbA1c, through episodic blood glucose readings, to continuous glucose monitoring (CGM) producing data points every few minutes. The high-temporal resolution of CGM data enabled increasingly intensive treatments, from decision support assisting insulin injection or oral medication, to automated closed-loop control, known as the "artificial pancreas." Throughout this progress, mathematical models and computer simulation of the human metabolic system became indispensable for the technological progress of diabetes treatment, enabling every step, from assessment of insulin sensitivity via the now classic Minimal Model of Glucose Kinetics, to in silico trials replacing animal experiments, to automated insulin delivery algorithms. In this review, we follow these developments, beginning with the Minimal Model, which evolved through the years to become large and comprehensive and trigger a paradigm change in the design of diabetes optimization strategies: in 2007, we introduced a sophisticated model of glucose-insulin dynamics and a computer simulator equipped with a "population" of N = 300 in silico "subjects" with type 1 diabetes. In January 2008, in an unprecedented decision, the Food and Drug Administration (FDA) accepted this simulator as a substitute to animal trials for the pre-clinical testing of insulin treatment strategies. This opened the field for rapid and cost-effective development and pre-clinical testing of new treatment approaches, which continues today. Meanwhile, animal experiments for the purpose of designing new insulin treatment algorithms have been abandoned.

Performance Effect of Adjusting Insulin Sensitivity for Model-Based Automated Insulin Delivery Systems

BACKGROUND

Model predictive control (MPC) has become one of the most popular control strategies for automated insulin delivery (AID) in type 1 diabetes (T1D). These algorithms rely on a prediction model to determine the best insulin dosing every sampling time. Although these algorithms have been shown to be safe and effective for glucose management through clinical trials, managing the ever-fluctuating relationship between insulin delivery and resulting glucose uptake (aka insulin sensitivity, IS) remains a challenge. We aim to evaluate the effect of informing an AID system with IS on the performance of the system.

METHOD

The University of Virginia (UVA) MPC control-based hybrid closed-loop (HCL) and fully closed-loop (FCL) system was used. One-day simulations at varying levels of IS were run with the UVA/Padova T1D Simulator. The AID system was informed with an estimated value of IS obtained through a mixed meal glucose tolerance test. Relevant controller parameters are updated to inform insulin dosing of IS. Performance of the HCL/FCL system with and without information of the changing IS was assessed using a novel performance metric penalizing the time outside the target glucose range.

RESULTS

Feedback in AID systems provides a certain degree tolerance to changes in IS. However, IS-informed bolus and basal dosing improve glycemic outcomes, providing increased protection against hyperglycemia and hypoglycemia according to the individual's physiological state.

CONCLUSIONS

The proof-of-concept analysis presented here shows the potentially beneficial effects on system performance of informing the AID system with accurate estimates of IS. In particular, when considering reduced IS, the informed controller provides increased protection against hyperglycemia compared with the naïve controller. Similarly, reduced hypoglycemia is obtained for situations with increased IS. Further tailoring of the adaptation schemes proposed in this work is needed to overcome the increased hypoglycemia observed in the more resistant cases and to optimize the performance of the adaptation method.

ReplayBG: A Digital Twin-Based Methodology to Identify a Personalized Model From Type 1 Diabetes Data and Simulate Glucose Concentrations to Assess Alternative Therapies

OBJECTIVE

Design and assessment of new therapies for type 1 diabetes (T1D) management can be greatly facilitated by in silico simulations. The ReplayBG simulation methodology here proposed allows "replaying" the scenario behind data already collected by simulating the glucose concentration obtained in response to alternative insulin/carbohydrate therapies and evaluate their efficacy leveraging the concept of digital twin.

METHODS

ReplayBG is based on two steps. First, a personalized model of glucose-insulin dynamics is identified using insulin, carbohydrate, and continuous glucose monitoring (CGM) data. Then, this model is used to simulate the glucose concentration that would have been obtained by "replaying" the same portion of data using a different therapy. The validity of the methodology was evaluated on 100 virtual subjects using the UVa/Padova T1D Simulator (T1DS). In particular, the glucose concentration traces simulated by ReplayBG are compared with those provided by T1DS in five different scenarios of insulin and carbohydrate treatment modifications. Furthermore, we compared ReplayBG with a state-of-the-art methodology for the scope. Finally, two case studies using real data are also presented.

RESULTS

ReplayBG simulates with high accuracy the effect of the considered insulin and carbohydrate treatment alterations, performing significantly better than state-of-art method in almost all considered situations.

CONCLUSION

ReplayBG proved to be a reliable and robust tool to retrospectively explore the effect of new treatments for T1D on the glucose dynamics. It is freely available as open source software at https://github.com/gcappon/replay-bg.

SIGNIFICANCE

ReplayBG offers a new approach to preliminary evaluate new therapies for T1D management before clinical trials.

A Markov Model of Gap Occurrence in Continuous Glucose Monitoring Data for Realistic in Silico Clinical Trials

BACKGROUND AND OBJECTIVE

Continuous glucose monitoring (CGM) sensors measure interstitial glucose concentration every 1-5 min for days or weeks. New CGM-based diabetes therapies are often tested in in silico clinical trials (ISCTs) using diabetes simulators. Accurate models of CGM sensor inaccuracies and failures could help improve the realism of ISCTs. However, the modeling of CGM failures has not yet been fully addressed in the literature. This work aims to develop a mathematical model of CGM gaps, i.e., occasional portions of missing data generated by temporary sensor errors (e.g., excessive noise or artifacts).

METHODS

Two datasets containing CGM traces collected in 167 adults and 205 children, respectively, using the Dexcom G6 sensor (Dexcom Inc., San Diego, CA) were used. Four Markov models, of increasing complexity, were designed to describe three main characteristics: number of gaps for each sensor, gap distribution in the monitoring days, and gap duration. Each model was identified on a portion of each dataset (training set). The remaining portion of each dataset (real test set) was used to evaluate model performance through a Monte Carlo simulation approach. Each model was used to generate 100 simulated test sets with the same size as the real test set. The distributions of gap characteristics on the simulated test sets were compared with those observed on the real test set, using the two-sample Kolmogorov-Smirnov test and the Jensen-Shannon divergence.

RESULTS

A six-state Markov model, having two states to describe normal sensor operation and four states to describe gap occurrence, achieved the best results. For this model, the Kolmogorov-Smirnov test found no significant differences between the distribution of simulated and real gap characteristics. Moreover, this model obtained significantly lower Jensen-Shannon divergence values than the other models.

CONCLUSIONS

A Markov model describing CGM gaps was developed and validated on two real datasets. The model describes well the number of gaps for each sensor, the gap distribution over monitoring days, and the gap durations. Such a model can be integrated into existing diabetes simulators to realistically simulate CGM gaps in ISCTs and thus enable the development of more effective and robust diabetes management strategies.

Controlling and powering a fully implantable artificial pancreas refillable by ingestible pills

Over the past decade, there has been a growing interest in the development of an artificial pancreas for intraperitoneal insulin delivery. Intraperitoneal implantable pumps guarantee more physiological glycemic control than subcutaneous wearable ones, for the treatment of type 1 diabetes. In this work, a fully implantable artificial pancreas refillable by ingestible pills is presented. In particular, solutions enabling the communication between the implanted pump and external user interfaces and novel control algorithms to intraperitoneally release an adequate amount of insulin based on glycemic data are shown. In addition, the powering and the wireless battery recharging are addressed. Specifically, the design and optimization of a customized transcutaneous energy transfer with two independent wireless channels are presented. The system was tested in terms of recharging efficacy, possible temperature rise within the body, during the recharging process and reliability of the wireless connection in the air and in the presence of ex vivo tissues.Clinical Relevance- This work aims to improve the control, battery recharging, and wireless communication of a fully implantable artificial pancreas for type 1 diabetes treatment.

In silico design and validation of a time-varying PID controller for an artificial pancreas with intraperitoneal insulin delivery and glucose sensing

Type 1 diabetes (T1D) is a chronic autoimmune disease featured by the loss of beta cell function and the need for lifetime insulin replacement. Over the recent decade, the use of automated insulin delivery systems (AID) has shifted the paradigm of treatment: the availability of continuous subcutaneous (SC) glucose sensors to guide SC insulin delivery through a control algorithm has allowed, for the first time, to reduce the daily burden of the disease as well as to abate the risk for hypoglycemia. AID use is still limited by individual acceptance, local availability, coverage, and expertise. A major drawback of SC insulin delivery is the need for meal announcement and the peripheral hyperinsulinemia that, over time, contributes to macrovascular complications. Inpatient trials using intraperitoneal (IP) insulin pumps have demonstrated that glycemic control can be improved without meal announcement due to the faster insulin delivery through the peritoneal space. This calls for novel control algorithms able to account for the specificities of IP insulin kinetics. Recently, our group described a two-compartment model of IP insulin kinetics demonstrating that the peritoneal space acts as a virtual compartment and IP insulin delivery is virtually intraportal (intrahepatic), thus closely mimicking the physiology of insulin secretion. The FDA-accepted T1D simulator for SC insulin delivery and sensing has been updated for IP insulin delivery and sensing. Herein, we design and validate-in silico-a time-varying proportional integrative derivative controller to guide IP insulin delivery in a fully closed-loop mode without meal announcement.

Minimally invasive electrochemical continuous glucose monitoring sensors: Recent progress and perspective

Diabetes and its complications are seriously threatening the health and well-being of hundreds of millions of people. Glucose levels are essential indicators of the health conditions of diabetics. Over the past decade, concerted efforts in various fields have led to significant advances in glucose monitoring technology. In particular, the rapid development of continuous glucose monitoring (CGM) based on electrochemical sensing principles has great potential to overcome the limitations of self-monitoring blood glucose (SMBG) in continuously tracking glucose trends, evaluating diabetes treatment options, and improving the quality of life of diabetics. However, the applications of minimally invasive electrochemical CGM sensors are still limited owing to the following aspects: i) invasiveness, ii) short lifespan, iii) biocompatibility, and iv) calibration and prediction. In recent years, the performance of minimally invasive electrochemical CGM systems (CGMSs) has been significantly improved owing to breakthrough developments in new materials and key technologies. In this review, we summarize the history of commercial CGMSs, the development of sensing principles, and the research progress of minimally invasive electrochemical CGM sensors in reducing the invasiveness of implanted probes, maintaining enzyme activity, and improving the biocompatibility of the sensor interface. In addition, this review also introduces calibration algorithms and prediction algorithms applied to CGMSs and describes the application of machine learning algorithms for glucose prediction.

Maximizing Glycemic Benefits of Using Faster Insulin Formulations in Type 1 Diabetes: In Silico Analysis Under Open- and Closed-Loop Conditions

Background: Ultrarapid-acting insulin analogs that could improve or even prevent postprandial hyperglycemia are now available for both research and clinical care. However, clear glycemic benefits remain elusive, especially when combined with automated insulin delivery (AID) systems. In this work, we study two insulin formulations in silico and highlight adjustments of both open-loop and closed-loop insulin delivery therapies as a critical step to achieve clinically meaningful improvements. Methods: Subcutaneous insulin transport models for two faster analogs, Fiasp (Novo Nordisk, Bagsværd, Denmark) and AT247 (Arecor, Saffron Walden, United Kingdom), were identified using data collected from prior clamp experiments, and integrated into the UVA/Padova type 1 diabetes simulator (adult cohort, N = 100). Pump therapy parameters and the aggressiveness of our full closed-loop algorithm were adapted to the new insulin pharmacokinetic and pharmacodynamic profiles through a sequence of in silico studies. Finally, we assessed these analogs' glycemic impact with and without modified therapy parameters in simulated conditions designed to match clinical trial data. Results: Simply switching to faster insulin analogs shows limited improvements in glycemic outcomes. However, when insulin acceleration is accompanied by therapy adaptation, clinical significance is found comparing time-in-range (70-180 mg/dL) with Aspart versus AT247 in open-loop (+5.1%); and Aspart versus Fiasp (+5.4%) or AT247 (+10.6%) in full closed-loop with no clinically significant differences in the exposure to hypoglycemia. Conclusion: In silico results suggest that properly adjusting intensive insulin therapy profiles, or AID tuning, to faster insulin analogs is necessary to obtain clinically significant improvements in glucose control.

Chinese diabetes datasets for data-driven machine learning

Data of the diabetes mellitus patients is essential in the study of diabetes management, especially when employing the data-driven machine learning methods into the management. To promote and facilitate the research in diabetes management, we have developed the ShanghaiT1DM and ShanghaiT2DM Datasets and made them publicly available for research purposes. This paper describes the datasets, which was acquired on Type 1 (n = 12) and Type 2 (n = 100) diabetic patients in Shanghai, China. The acquisition has been made in real-life conditions. The datasets contain the clinical characteristics, laboratory measurements and medications of the patients. Moreover, the continuous glucose monitoring readings with 3 to 14 days as a period together with the daily dietary information are also provided. The datasets can contribute to the development of data-driven algorithms/models and diabetes monitoring/managing technologies.

Evaluation of a Web-Based Simulation Tool for Self-Management Support in Type 1 Diabetes: A Pilot Study

OBJECTIVE

To develop and evaluate a novel Web-based Simulation Tool (WST) that brings simulation technologies to people with Type 1 Diabetes (T1D), enabling unique patient-data interactions seamlessly on a daily basis.

METHODS

A pilot clinical trial was conducted to assess system usability. The study consisted of one week of observation (Phase 1) and four weeks of interaction with WST (Phase 2). Responses to Technology Acceptance (TA) and Diabetes Distress Scale (DDS) questionnaires were collected, and follow-up interviews were conducted after Phase 2.

RESULTS

Fifteen participants with T1D using Control-IQ technology (age: 36 ± 13 years, HbA1c: 6.5% ± 0.7%) completed all study procedures. Generated simulation models achieved a median Mean Absolute Relative Difference (MARD) of 6.8% [interquartile range, IQR: 5.1%, 9.1%]. A decrease in expected benefits (likely explained by issues with the third-party data collection system) and an increase in expected burdens were observed. On a 1-5 scale, ease of use, trust, and usefulness scores were 3 [3,4], 4 [3,4], and 4 [3,4], respectively. Time below 70 mg/dL decreased between Phases 1 and 2 (1.6% [0.7%,3.7%] vs 0.8% [0.5%,3.0%]). A reduction in mean emotional burden was also observed (2.5 ± 1.1 vs 2.1 ± 0.8).

CONCLUSIONS

Results indicate that there was a learning curve to WST, but also that most participants trusted the system and found it useful in their diabetes care.

SIGNIFICANCE

Simulation technologies like WST could be used by educators and patients to facilitate diabetes self-management, leading to better diabetes literacy and reducing associated distress.

A modelling approach to hepatic glucose production estimation

Stable isotopes are currently used to measure glucose fluxes responsible for observed glucose concentrations, providing information on hepatic and peripheral insulin sensitivity. The determination of glucose turnover, along with fasting and postprandial glucose concentrations, is relevant for inferring insulin sensitivity levels. At equilibrium (e.g. during the fasting state) the rate of glucose entering the circulation equals its rate of disappearance from the circulation. If under these conditions tracer is infused at a constant rate and Specific Activity (SA) or Tracer to Tracee (TTR) ratio is computed, the Rate of Appearance (RA) equals the Rate of Disappearance (RD) and equals the ratio between infusion rate and TTR or SA. In the post-prandial situation or during perturbation studies, however, estimation of RA and RD becomes more complex because they are not necessarily equal and, furthermore, may vary over time due to gastric emptying, glucose absorption, appearance of ingested or infused glucose, variations of EGP and glucose disappearance. Up to now, the most commonly used approach to compute RA, RD and EGP has been the single-pool model by Steele. Several authors, however, report pitfalls in the use of this method, such as "paradoxical" increase in EGP immediately after meal ingestion and "negative" rates of EGP. Different attempts have been made to reduce the impact of these errors, but the same problems are still encountered. In the present work a completely different approach is proposed, where cold and labeled [6, 6-2H2] glucose observations are simultaneously fitted and where both RD and EGP are represented by simple but reasonable functions. As an example, this approach is applied to an intra-venous experiment, where cold glucose is infused at variable rates to reproduce a desired glycaemic time-course. The goal of the present work is to show that appropriate, if simple, modelling of the whole infusion procedure together with the underlying physiological system allows robust estimation of EGP with single-tracer administration, without the artefacts produced by the Steele method.

Distributed Assessment of Virtual Insulin-Pump Settings Using SmartCGMS and DMMS.R for Diabetes Treatment

Diabetes is a heterogeneous group of diseases that share a common trait of elevated blood glucose levels. Insulin lowers this level by promoting glucose utilization, thus avoiding short- and long-term organ damage due to the elevated blood glucose level. A patient with diabetes uses an insulin pump to dose insulin. The pump uses a controller to compute and dose the correct amount of insulin to keep blood glucose levels in a safe range. Insulin-pump controller development is an ongoing process aiming at fully closed-loop control. Controllers entering the market must be evaluated for safety. We propose an evaluation method that exploits an FDA-approved diabetic patient simulator. The method evaluates a Cartesian product of individual insulin-pump parameters with a fine degree of granularity. As this is a computationally intensive task, the simulator executes on a distributed cluster. We identify safe and risky combinations of insulin-pump parameter settings by applying the binomial model and decision tree to this product. As a result, we obtain a tool for insulin-pump settings and controller safety assessment. In this paper, we demonstrate the tool with the Low-Glucose Suspend and OpenAPS controllers. For average ± standard deviation, LGS and OpenAPS exhibited 1.7 ± 0.6% and 3.2 ± 1.8% of local extrema (i.e., good insulin-pump settings) out of all the entire Cartesian products, respectively. A continuous region around the best-discovered settings (i.e., the global extremum) of the insulin-pump settings spread across 4.0 ± 1.1% and 4.1 ± 1.3% of the Cartesian products, respectively.

Impact of Carbohydrate Counting Error on Glycemic Control in Open-Loop Management of Type 1 Diabetes: Quantitative Assessment Through an In Silico Trial

BACKGROUND

In the management of type 1 diabetes (T1D), systematic and random errors in carb-counting can have an adverse effect on glycemic control. In this study, we performed an in silico trial aiming at quantifying the impact of different levels of carb-counting error on glycemic control.

METHODS

The T1D patient decision simulator was used to simulate 7-day glycemic profiles of 100 adults using open-loop therapy. The simulation was repeated for different values of systematic and random carb-counting errors, generated with Gaussian distribution varying the error mean from -10% to +10% and standard deviation (SD) from 0% to 50%. The effect of the error was evaluated by computing the difference of time inside (∆TIR), above (∆TAR) and below (∆TBR) the target glycemic range (70-180mg/dl) compared to the reference case, that is, absence of error. Finally, 3 linear regression models were developed to mathematically describe how error mean and SD variations result in ∆TIR, ∆TAR, and ∆TBR changes.

RESULTS

Random errors globally deteriorate the glycemic control; systematic underestimations lead to, on average, up to 5.2% more TAR than the reference case, while systematic overestimation results in up to 0.8% more TBR. The different time in range metrics were linearly related with error mean and SD (R²>0.95), with slopes of <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"><mml:mrow><mml:msub><mml:mi>β</mml:mi><mml:mrow><mml:mi>M</mml:mi><mml:mi>E</mml:mi><mml:mi>A</mml:mi><mml:mi>N</mml:mi></mml:mrow></mml:msub><mml:mo>=</mml:mo><mml:mn>0</mml:mn><mml:mo>.</mml:mo><mml:mn>21</mml:mn></mml:mrow></mml:math>, <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"><mml:mrow><mml:msub><mml:mi>β</mml:mi><mml:mrow><mml:mi>S</mml:mi><mml:mi>D</mml:mi></mml:mrow></mml:msub><mml:mo>=</mml:mo><mml:mo>-</mml:mo><mml:mn>0</mml:mn><mml:mo>.</mml:mo><mml:mn>07</mml:mn></mml:mrow></mml:math> for ∆TIR, <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"><mml:mrow><mml:msub><mml:mi>β</mml:mi><mml:mrow><mml:mi>M</mml:mi><mml:mi>E</mml:mi><mml:mi>A</mml:mi><mml:mi>N</mml:mi></mml:mrow></mml:msub><mml:mo>=</mml:mo><mml:mo>-</mml:mo><mml:mn>0</mml:mn><mml:mo>.</mml:mo><mml:mn>25</mml:mn></mml:mrow></mml:math>, <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"><mml:mrow><mml:msub><mml:mi>β</mml:mi><mml:mrow><mml:mi>S</mml:mi><mml:mi>D</mml:mi></mml:mrow></mml:msub><mml:mo>=</mml:mo><mml:mo>+</mml:mo><mml:mn>0</mml:mn><mml:mo>.</mml:mo><mml:mn>06</mml:mn></mml:mrow></mml:math> for ∆TAR, and <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"><mml:mrow><mml:msub><mml:mi>β</mml:mi><mml:mrow><mml:mi>M</mml:mi><mml:mi>E</mml:mi><mml:mi>A</mml:mi><mml:mi>N</mml:mi></mml:mrow></mml:msub><mml:mo>=</mml:mo><mml:mn>0</mml:mn><mml:mo>.</mml:mo><mml:mn>05</mml:mn></mml:mrow></mml:math>, <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"><mml:mrow><mml:msub><mml:mi>β</mml:mi><mml:mrow><mml:mi>S</mml:mi><mml:mi>D</mml:mi></mml:mrow></mml:msub><mml:mo>=</mml:mo><mml:mo>+</mml:mo><mml:mn>0</mml:mn><mml:mo>.</mml:mo><mml:mn>01</mml:mn></mml:mrow></mml:math> for ∆TBR.

CONCLUSIONS

The quantification of carb-counting error impact performed in this work may be useful understanding causes of glycemic variability and the impact of possible therapy adjustments or behavior changes in different glucose metrics.

A physical activity-intensity driven glycemic model for type 1 diabetes

BACKGROUND AND OBJECTIVE

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.

METHODS

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.

RESULTS

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).

CONCLUSIONS

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.

Gluclas: A software for computer-aided modulation of glucose infusion in glucose clamp experiments

BACKGROUND AND OBJECTIVE

The glucose clamp (GC) is an experimental technique for assessing several aspects of glucose metabolism. In these experiments, investigators face the non-trivial challenge of accurately adjusting the rate of intravenous glucose infusion to drive subjects' blood glucose (BG) concentration towards a desired plateau level. In this work we present Gluclas, an open-source software to support researchers in the modulation of glucose infusion rate (GIR) during GC experiments.

METHODS

Gluclas uses a proportional-integrative-derivative controller to provide GIR suggestions based on BG measurements. The controller embeds an anti-wind-up scheme to account for actuator physical limits and suitable corrections of control action to accommodate for possible sampling jitter due to manual measurement and actuation. The software also provides a graphic user interface to increase its usability. A preliminary validation of the controller is performed for different clamp scenarios (hyperglycemic, euglycemic, hypoglycemic) on a simulator of glucose metabolism in healthy subjects, which also includes models of measurement error and sampling delay for increased realism. In silico trials are performed on 50 virtual subjects. We also report the results of the first in-vivo application of the software in three subjects undergoing a hypoglycemic clamp.

RESULTS

In silico, during the plateau period, the coefficient of variation (CV) is in median below 5% for every protocol, with 5% being considered the threshold for sufficient quality. In terms of median [5th percentile, 95th percentile], average BG level during the plateau period is 12.18 [11.58 - 12.53] mmol/l in the hyperglycemic clamp (target: 12.4 mmol/), 4.92 [4.51 - 5.14] mmol/l in the euglycemic clamp (target: 5.5 mmol/) and 2.38 [2.33 - 2.64] in the hypoglycemic clamp (target: 2.5 mmol/). Results in vivo are consistent with those obtained in silico during the plateau period: average BG levels are between 2.56 and 2.68 mmol/l (target: 2.5 mmol/l); CV is below 5% for all three experiments.

CONCLUSIONS

Gluclas offered satisfactory control for tested GC protocols. Although its safety and efficacy need to be further validated in vivo, this preliminary validation suggest that Gluclas offers a reliable and non-expensive solution for reducing investigator bias and improving the quality of GC experiments.

Evaluation of blood glucose level control in type 1 diabetic patients using deep reinforcement learning

Diabetes mellitus is a disease associated with abnormally high levels of blood glucose due to a lack of insulin. Combining an insulin pump and continuous glucose monitor with a control algorithm to deliver insulin is an alternative to patient self-management of insulin doses to control blood glucose levels in diabetes mellitus patients. In this work, we propose a closed-loop control for blood glucose levels based on deep reinforcement learning. We describe the initial evaluation of several alternatives conducted on a realistic simulator of the glucoregulatory system and propose a particular implementation strategy based on reducing the frequency of the observations and rewards passed to the agent, and using a simple reward function. We train agents with that strategy for three groups of patient classes, evaluate and compare it with alternative control baselines. Our results show that our method is able to outperform baselines as well as similar recent proposals, by achieving longer periods of safe glycemic state and low risk.

Minimal and Maximal Models to Quantitate Glucose Metabolism: Tools to Measure, to Simulate and to Run in Silico Clinical Trials

Several models have been proposed to describe the glucose system at whole-body, organ/tissue and cellular level, designed to measure non-accessible parameters (minimal models), to simulate system behavior and run in silico clinical trials (maximal models). Here, we will review the authors' work, by putting it into a concise historical background. We will discuss first the parametric portrait provided by the oral minimal models-building on the classical intravenous glucose tolerance test minimal models-to measure otherwise non-accessible key parameters like insulin sensitivity and beta-cell responsivity from a physiological oral test, the mixed meal or the oral glucose tolerance tests, and what can be gained by adding a tracer to the oral glucose dose. These models were used in various pathophysiological studies, which we will briefly review. A deeper understanding of insulin sensitivity can be gained by measuring insulin action in the skeletal muscle. This requires the use of isotopic tracers: both the classical multiple-tracer dilution and the positron emission tomography techniques are discussed, which quantitate the effect of insulin on the individual steps of glucose metabolism, that is, bidirectional transport plasma-interstitium, and phosphorylation. Finally, we will present a cellular model of insulin secretion that, using a multiscale modeling approach, highlights the relations between minimal model indices and subcellular secretory events. In terms of maximal models, we will move from a parametric to a flux portrait of the system by discussing the triple tracer meal protocol implemented with the tracer-to-tracee clamp technique. This allows to arrive at quasi-model independent measurement of glucose rate of appearance (Ra), endogenous glucose production (EGP), and glucose rate of disappearance (Rd). Both the fast absorbing simple carbs and the slow absorbing complex carbs are discussed. This rich data base has allowed us to build the UVA/Padova Type 1 diabetes and the Padova Type 2 diabetes large scale simulators. In particular, the UVA/Padova Type 1 simulator proved to be a very useful tool to safely and effectively test in silico closed-loop control algorithms for an artificial pancreas (AP). This was the first and unique simulator of the glucose system accepted by the U.S. Food and Drug Administration as a substitute to animal trials for in silico testing AP algorithms. Recent uses of the simulator have looked at glucose sensors for non-adjunctive use and new insulin molecules.

A software interface for in silico testing of type 2 diabetes treatments

BACKGROUND AND OBJECTIVE

The increasing incidence of diabetes continuously stimulates the research on new antidiabetic drugs. Computer simulation can save time and costs, alleviating the need of animal trials and providing useful information for optimal experiment design and drug dosing. We recently presented a type 2 diabetes (T2D) simulator as tool for in silico testing of new molecules and guiding treatment optimization. Here we present a user-friendly interface aimed to increase the usability of the simulator.

METHOD

The simulator, based on a large-scale glucose, insulin, and C-peptide model and equipped with 100 virtual subjects well describing system dynamics in a real T2D population, is extended to incorporate pharmacokinetics/pharmacodynamics (PK/PD) of a drug of interest. A graphical interface is developed on top of the simulator, allowing an easy design of in silico experiments: specifically, it is possible to select the population size to test, design the experiment (crossover or parallel), its duration and the sampling grid, choose glucose and insulin doses, and define treatment PK/PD and dose administered. The simulator also provides the outcome metrics requested by the user, and performs statistical comparisons among treatments and/or placebo.

RESULTS

To illustrate the potential of the simulator, we provided a case study using metformin and liraglutide. Literature-based PK/PD models of metformin and liraglutide have been incorporated in the simulator, by modulating key drug-sensitive model parameters. An in silico placebo-controlled trial has been done by simulating a three-arm meal tolerance test with subjects receiving placebo, metformin 850 mg, liraglutide 1.80 mg, respectively. The obtained results are in agreement with the clinical evidences, in terms of main glucose, insulin, and C-peptide outcome metrics.

CONCLUSIONS

We developed a user-friendly software interface for the T2D simulator to support the design and test of new antidiabetic drugs and treatments. This increases the simulator usability, making it suitable also for users who have low experience with computer programming.

A simulator with realistic and challenging scenarios for virtual T1D patients undergoing CSII and MDI therapy

In silico simulations have become essential for the development of diabetes treatments. However, currently available simulators are not challenging enough and often suffer from limitations in insulin and meal absorption variability, which is unable to realistically reflect the dynamics of people with type 1 diabetes (T1D). Additionally, T1D simulators are mainly designed for the testing of continuous subcutaneous insulin infusion (CSII) therapies. In this work, a simulator is presented that includes a generated virtual patient (VP) cohort and both fast- and long-acting Glargine-100 U/ml (Gla-100), Glargine-300 U/ml (Gla-300), and Degludec-100 U/ml (Deg-100) insulin models. Therefore, in addition to CSII therapies, multiple daily injections (MDI) therapies can also be tested. The Hovorka model and its published parameter probability distributions were used to generate cohorts of VPs that represent a T1D population. Valid patients are filtered through restrictions that guarantee that they are physiologically acceptable. To obtain more realistic scenarios, basal insulin profile patterns from the literature have been used to identify variability in insulin sensitivity. A library of mixed meals identified from real data has also been included. This work presents and validates a methodology for the creation of realistic VP cohorts that include physiological variability and a simulator that includes challenging and realistic scenarios for in silico testing. A cohort of 47 VPs has been generated and in silico simulations of both CSII and MDI therapies were performed in open-loop. The simulation outcome metrics were contrasted with literature results.

A Reinforcement Learning Based System for Blood Glucose Control without Carbohydrate Estimation in Type 1 Diabetes: In Silico Validation

Type 1 Diabetes (T1D) is a chronic autoimmune disease, which requires the use of exogenous insulin for glucose regulation. In current hybrid closed-loop systems, meal entry is manual which adds cognitive burden to the persons living with T1D. In this study, we proposed a control system based on Proximal Policy Optimisation (PPO) that controls both basal and bolus insulin infusion and only requires meal announcement, thus eliminating the need for carbohydrate estimation. We evaluated the system on a challenging meal scenario, using an open-source simulator based on the UVA/Padova 2008 model and achieved a mean Time in Range value of 65% for the adult subject cohort, while maintaining a moderate hypoglycemic and hyperglycemic risk profile. The approach shows promise and welcomes further research towards the translation to a real-life artificial pancreas. Clinical relevance- This was an in-silico analysis towards the development of an autonomous artificial pancreas system for glucose control. The proposed system show promise in eliminating the need for estimating the carbohydrate content in meals.

Data-Driven Supervised Compression Artifacts Detection on Continuous Glucose Sensors

Continuous Glucose Monitoring (CGM) sensors micro-invasively provide frequent glucose readings, improving the management of Type 1 diabetic patients' life and making available reach data-sets for retrospective analysis. Unlikely, CGM sensors are subject to failures, such as compression artifacts, that might impact on both real-time and respective CGM use. In this work is focused on retrospective detection of compression artifacts. An in-silico dataset is generated using the T1D UVa/Padova simulator and compression artifacts are subsequently added in known position, thus creating a dataset with perfectly accurate faulty/not-faulty labels. The problem of compression artifact detection is then faced with supervised data-driven techniques, in particular using Random Forest algorithm. The detection performance guaranteed by the method on in-silico data is satisfactory, opening the way for further analysis on real-data.

A Dual-Hormone Multicontroller for Artificial Pancreas Systems

Artificial pancreas (AP) algorithms can be divided into single-hormone (SH) and dual-hormone (DH). SH algorithms regulate glycemia using insulin as their control input. On the other hand, DH algorithms also use glucagon to counteract insulin. While SH-AP systems are already commercially available, DH-AP systems are still in an earlier research phase. DH-AP systems have been questioned since the added complexity of glucagon infusion does not always guarantee hypoglycemia prevention and might significantly raise insulin delivery. In this work, a DH multicontroller is proposed based on a SH linear quadratic gaussian (LQG) algorithm with an additional LQG controller to deliver glucagon. This strategy has a switched structure that allows activating one of the following three controllers when necessary: a conservative insulin LQG controller to modulate basal delivery ( K₁), an aggressive insulin LQG controller to counteract meals ( K₂), or a glucagon LQG controller to avoid imminent hypoglycemia ( K₃). Here, an in silico study of the benefits of incorporating controller K₃ is carried out. Intra-patient variability and mixed meals are considered. Results indicate that the proposed switched, dual-hormone strategy yields to a reduction in hypoglycemia without increasing hyperglycemia, with no significant rise in insulin delivery.

Generation of post-meal insulin correction boluses in type 1 diabetes simulation models for in-silico clinical trials: More realistic scenarios obtained using a decision tree approach

BACKGROUND AND OBJECTIVE

In type 1 diabetes (T1D) research, in-silico clinical trials (ISCTs) notably facilitate the design/testing of new therapies. Published simulation tools embed mathematical models of blood glucose (BG) and insulin dynamics, continuous glucose monitoring (CGM) sensors, and insulin treatments, but lack a realistic description of some aspects of patient lifestyle impacting on glucose control. Specifically, to effectively simulate insulin correction boluses, required to treat post-meal hyperglycemia (BG > 180 mg/dL), the timing of the bolus may be influenced by subjects' behavioral attitudes. In this work, we develop an easily interpretable model of the variability of correction bolus timing observed in real data, and embed it into a popular simulation tool for ISCTs.

METHODS

Using data collected in 196 adults with T1D monitored in free-living conditions, we trained a decision tree (DT) model to classify whether a correction bolus is injected in a future time window, based on predictors collected back in time, related to CGM data, previous insulin boluses and subject's characteristics. The performance was compared to that of a logistic regression classifier with LASSO regularization (LC), trained on the same dataset. After validation, the DT was embedded within a popular T1D simulation tool and an ISCT was performed to compare the simulated correction boluses against those observed in a subset of data not used for model training.

RESULTS

The DT provided better classification performance (accuracy: 0.792, sensitivity: 0.430, specificity: 0.878, precision: 0.455) than the LC and presented good interpretability. The most predictive features were related to CGM (and its temporal variations), time since the last insulin bolus, and time of the day. The correction boluses simulated by the DT, after implementation in the simulation tool, showed a good agreement with real-world data.

CONCLUSIONS

The DT developed in this work represents a simple set of rules to mimic the same timing of correction boluses observed on real data. The inclusion of the model in simulation tools allows investigators to perform ISCTs that more realistically represent the patient behavior in taking correction boluses and the post-prandial BG response. In the future, more complex models can be investigated.

A metamodel-based flexible insulin therapy for type 1 diabetes patients subjected to aerobic physical activity

Patients with type 1 diabetes are subject to exogenous insulin injections, whether manually or through (semi)automated insulin pumps. Basic knowledge of the patient's characteristics and flexible insulin therapy (FIT) parameters are then needed. Specifically, artificial pancreas-like closed-loop insulin delivery systems are some of the most promising devices for substituting for endogenous insulin secretion in type 1 diabetes patients. However, these devices require self-reported information such as carbohydrates or physical activity from the patient, introducing potential miscalculations and delays that can have life-threatening consequences. Here, we display a metamodel for glucose-insulin dynamics that is subject to carbohydrate ingestion and aerobic physical activity. This metamodel incorporates major existing knowledge-based models. We derive comprehensive and universal definitions of the underlying FIT parameters to form an insulin sensitivity factor (ISF). In addition, the relevance of physical activity modelling is assessed, and the FIT is updated to take physical exercise into account. Specifically, we cope with physical activity by using heart rate sensors (watches) with a fully automated closed insulin loop, aiming to maximize the time spent in the glycaemic range (75.5% in the range and 1.3% below the range for hypoglycaemia on a virtual patient simulator).These mathematical parameter definitions are interesting on their own, may be new tools for assessing mathematical models and can ultimately be used in closed-loop artificial pancreas algorithms or to extend distinguished FIT.

Prediction of Glucose Concentration in Children with Type 1 Diabetes Using Neural Networks: An Edge Computing Application

Background: Type 1 Diabetes Mellitus (T1D) is an autoimmune disease that can cause serious complications that can be avoided by preventing the glycemic levels from exceeding the physiological range. Straightforwardly, many data-driven models were developed to forecast future glycemic levels and to allow patients to avoid adverse events. Most models are tuned on data of adult patients, whereas the prediction of glycemic levels of pediatric patients has been rarely investigated, as they represent the most challenging T1D population. Methods: A Convolutional Neural Network (CNN) and a Long Short-Term Memory (LSTM) Recurrent Neural Network were optimized on glucose, insulin, and meal data of 10 virtual pediatric patients. The trained models were then implemented on two edge-computing boards to evaluate the feasibility of an edge system for glucose forecasting in terms of prediction accuracy and inference time. Results: The LSTM model achieved the best numeric and clinical accuracy when tested in the .tflite format, whereas the CNN achieved the best clinical accuracy in uint8. The inference time for each prediction was far under the limit represented by the sampling period. Conclusion: Both models effectively predict glucose in pediatric patients in terms of numerical and clinical accuracy. The edge implementation did not show a significant performance decrease, and the inference time was largely adequate for a real-time application.

Two Decades of Research in Artificial Pancreas: Insights from a Bibliometric Analysis

BACKGROUND

Artificial pancreas is a well-known research topic devoted to achieving better glycemic outcomes that has been attracting increasing attention over the years. However, there is a lack of systematic, chronological, and synthesizing studies that show the background of the knowledge generation in this field. This study implements a bibliometric analysis to recognize the main documents, type of publications, research categories, countries, keywords, organizations, and authors related to this topic.

METHODS

Web of Science core collection database was accessed from 2000 to 2020 in order to select high-quality scientific documents based on a specific search query. Bibexcel, MS Excel, Power BI, R-Studio, VOSviewer, and CorText software were used for a descriptive and network analysis based on the local database obtained. Bibliometric parameters as the h-index, frequencies, co-authorship and co-ocurrences were computed.

RESULTS

A total of 756 documents were included that show a growing scientific production on this topic with an increasing contribution from engineering. Outstanding authors, organizations, and countries were identified. An analysis of trends in research was conducted according to the scientific categories of the Web of Science database to identify the main research interests of the last 2 decades and the emerging areas with greater prominence in the coming years. A keyword network analysis allowed to identify the main stages in the development of the AP research over time.

CONCLUSIONS

Results reveal a comprehensive background of the knowledge generation for the AP topic during the last 2 decades, which has been strengthened with international collaborations and a remarkable interdisciplinarity between endocrinology and engineering, giving rise to a growing number of research areas over time, where computer science and medical informatics stand out as the main emerging research areas.

Modeling the Effect of Subcutaneous Lixisenatide on Glucoregulatory Endocrine Secretions and Gastric Emptying in Type 2 Diabetes to Simulate the Effect of iGlarLixi Administration Timing on Blood Sugar Profiles

BACKGROUND

As type 2 diabetes (T2D) progresses, intensification to combination therapies, such as iGlarLixi (a fixed-ratio GLP-1 RA and basal insulin combination), may be required. Here a simulation study was used to assess the effect of iGlarLixi administration timing (am vs pm) on blood sugar profiles.

METHODS

Models of lixisenatide were built with a selection procedure, optimizing measurement fits and model complexity, and were included in a pre-existing T2D simulation platform containing glargine models. With the resulting tool, a simulated trial was conducted with 100 in-silico participants with T2D. Individuals were given iGLarLixi either before breakfast or before an evening meal for 2 weeks and daily glycemic profiles were analyzed. In the model, breakfast was considered the largest meal of the day.

RESULTS

A similar percentage of time within 24 hours was spent with blood sugar levels between 70 to 180 mg/dL when iGlarLixi was administered pre-breakfast or pre-evening meal (73% vs 71%, respectively). Overall percent of time with blood glucose levels above 180 mg/dL within a 24-hour period was similar when iGlarLixi was administered pre-breakfast or pre-evening meal (26% vs 28%, respectively). Rates of hypoglycemia were low in both regimens, with a blood glucose concentration of below 70 mg/dL only observed for 1% of the 24-hour time period for either timing of administration.

CONCLUSIONS

Good efficacy was observed when iGlarlixi was administered pre-breakfast; however, administration of iGlarlixi pre-evening meal was also deemed to be effective, even though in the model the size of the evening meal was smaller than that of the breakfast.

Integrating Multiple Inputs Into an Artificial Pancreas System: Narrative Literature Review

BACKGROUND

Type 1 diabetes (T1D) is a chronic autoimmune disease in which a deficiency in insulin production impairs the glucose homeostasis of the body. Continuous subcutaneous infusion of insulin is a commonly used treatment method. Artificial pancreas systems (APS) use continuous glucose level monitoring and continuous subcutaneous infusion of insulin in a closed-loop mode incorporating a controller (or control algorithm). However, the operation of APS is challenging because of complexities arising during meals, exercise, stress, sleep, illnesses, glucose sensing and insulin action delays, and the cognitive burden. To overcome these challenges, options to augment APS through integration of additional inputs, creating multi-input APS (MAPS), are being investigated.

OBJECTIVE

The aim of this survey is to identify and analyze input data, control architectures, and validation methods of MAPS to better understand the complexities and current state of such systems. This is expected to be valuable in developing improved systems to enhance the quality of life of people with T1D.

METHODS

A literature survey was conducted using the Scopus, PubMed, and IEEE Xplore databases for the period January 1, 2005, to February 10, 2020. On the basis of the search criteria, 1092 articles were initially shortlisted, of which 11 (1.01%) were selected for an in-depth narrative analysis. In addition, 6 clinical studies associated with the selected studies were also analyzed.

RESULTS

Signals such as heart rate, accelerometer readings, energy expenditure, and galvanic skin response captured by wearable devices were the most frequently used additional inputs. The use of invasive (blood or other body fluid analytes) inputs such as lactate and adrenaline were also simulated. These inputs were incorporated to switch the mode of the controller through activity detection, directly incorporated for decision-making and for the development of intermediate modules for the controller. The validation of the MAPS was carried out through the use of simulators based on different physiological models and clinical trials.

CONCLUSIONS

The integration of additional physiological signals with continuous glucose level monitoring has the potential to optimize glucose control in people with T1D through addressing the identified limitations of APS. Most of the identified additional inputs are related to wearable devices. The rapid growth in wearable technologies can be seen as a key motivator regarding MAPS. However, it is important to further evaluate the practical complexities and psychosocial aspects associated with such systems in real life.

Intraperitoneal Insulin Delivery: Evidence of a Physiological Route for Artificial Pancreas From Compartmental Modeling

BACKGROUND

Intraperitoneal insulin delivery has proven to safely overcome a major limit of subcutaneous delivery-meal announcement-and has been able to optimize glycemic control in adults under controlled experimental conditions. In addition, intraperitoneal delivery avoids peripheral hyperinsulinemia resulting from the subcutaneous route and restores a physiological liver gradient.

METHODS

Relying on a unique data set of intraperitoneal closed-loop insulin delivery obtained with a Model Predictive Controller (MPC), we develop a compartmental model of intraperitoneal insulin kinetics, which, once included in the UVa/Padova T1D simulator, will facilitate the investigation of various control strategies, for example, the simpler Proportional Integral Derivative controller versus MPC.

RESULTS

Intraperitoneal insulin kinetics can be described with a 2-compartment model including liver and plasma.

CONCLUSION

Intraperitoneal insulin transit is fast enough to render irrelevant the addition of a peritoneal compartment, proving the peritoneum being a virtual-not actual-transit space for insulin delivery.

Dynamic of Glucose Homeostasis in Virtual Patients: A Comparison between Different Behaviors

This work presents a mathematical model of homeostasis dynamics in healthy individuals, focusing on the generation of conductive data on glucose homeostasis throughout the day under dietary and physical activity factors. Two case studies on glucose dynamics for populations under conditions of physical activity and sedentary lifestyle were developed. For this purpose, two types of virtual populations were generated, the first population was developed according to the data of a total of 89 physical persons between 20 and 75 years old and the second was developed using the Monte Carlo approach, obtaining a total of 200 virtual patients. In both populations, each participant was classified as an active or sedentary person depending on the physical activity performed. The results obtained demonstrate the capacity of virtual populations in the generation of in-silico approximations similar to those obtained from in-vivo studies. Obtaining information that is only achievable through specific in-vivo experiments. Being a tool that generates information for the approach of alternatives in the prevention of the development of type 2 Diabetes.

Towards the Integration of an Islet-Based Biosensor in Closed-Loop Therapies for Patients With Type 1 Diabetes

In diabetes mellitus (DM) treatment, Continuous Glucose Monitoring (CGM) linked with insulin delivery becomes the main strategy to improve therapeutic outcomes and quality of patients' lives. However, Blood Glucose (BG) regulation with CGM is still hampered by limitations of algorithms and glucose sensors. Regarding sensor technology, current electrochemical glucose sensors do not capture the full spectrum of other physiological signals, i.e., lipids, amino acids or hormones, relaying the general body status. Regarding algorithms, variability between and within patients remains the main challenge for optimal BG regulation in closed-loop therapies. This work highlights the simulation benefits to test new sensing and control paradigms which address the previous shortcomings for Type 1 Diabetes (T1D) closed-loop therapies. The UVA/Padova T1DM Simulator is the core element here, which is a computer model of the human metabolic system based on glucose-insulin dynamics in T1D patients. That simulator is approved by the US Food and Drug Administration (FDA) as an alternative for pre-clinical testing of new devices and closed-loop algorithms. To overcome the limitation of standard glucose sensors, the concept of an islet-based biosensor, which could integrate multiple physiological signals through electrical activity measurement, is assessed here in a closed-loop insulin therapy. This investigation has been addressed by an interdisciplinary consortium, from endocrinology to biology, electrophysiology, bio-electronics and control theory. In parallel to the development of an islet-based closed-loop, it also investigates the benefits of robust control theory against the natural variability within a patient population. Using 4 meal scenarios, numerous simulation campaigns were conducted. The analysis of their results then introduces a discussion on the potential benefits of an Artificial Pancreas (AP) system associating the islet-based biosensor with robust algorithms.

Using an Online Disturbance Rejection and Anticipation System to Reduce Hyperglycemia in a Fully Closed-Loop Artificial Pancreas System

INTRODUCTION

Hyperglycemia following meals is a recurring challenge for people with type 1 diabetes, and even the most advanced available automated systems currently require manual input of carbohydrate amounts. To progress toward fully automated systems, we present a novel control system that can automatically deliver priming boluses and/or anticipate eating behaviors to improve postprandial full closed-loop control.

METHODS

A model predictive control (MPC) system was enhanced by an automated bolus system reacting to early glucose rise and/or a multistage MPC (MS-MPC) framework to anticipate historical patterns. Priming was achieved by detecting large glycemic disturbances, such as meals, and delivering a fraction of the patient's total daily insulin (TDI) modulated by the disturbance's likelihood (bolus priming system [BPS]). In the anticipatory module, glycemic disturbance profiles were generated from historical data using clustering to group days with similar behaviors; the probability of each cluster is then evaluated at every controller step and informs the MS-MPC framework to anticipate each profile. We tested four configurations: MPC, MPC + BPS, MS-MPC, and MS-MPC + BPS in simulation to contrast the effect of each controller module.

RESULTS

Postprandial time in range was highest for MS-MPC + BPS: 60.73 ± 25.39%, but improved with each module: MPC + BPS: 56.95±25.83 and MS-MPC: 54.83 ± 26.00%, compared with MPC: 51.79 ± 26.12%. Exposure to hypoglycemia was maintained for all controllers (time below 70 mg/dL <0.5%), and improvement came primarily from a reduction in postprandial time above range (MS-MPC + BPS: 39.10 ± 25.32%, MPC + BPS: 42.99 ± 25.81%, MS-MPC: 45.09 ± 25.96%, MPC: 48.18 ± 26.09%).

CONCLUSIONS

The BPS and anticipatory disturbance profiles improved blood glucose control and were most efficient when combined.