Advanced Insulin Bolus Advisor Based on Run-To-Run Control and Case-Based Reasoning

Design and In Silico Evaluation of an Exercise Decision Support System Using Digital Twin Models

BACKGROUND

Managing glucose levels during exercise is challenging for individuals with type 1 diabetes (T1D) since multiple factors including activity type, duration, intensity and other factors must be considered. Current decision support tools lack personalized recommendations and fail to distinguish between aerobic and resistance exercise. We propose an exercise-aware decision support system (exDSS) that uses digital twins to deliver personalized recommendations to help people with T1D maintain safe glucose levels (70-180 mg/dL) and avoid low glucose (<70 mg/dL) during and after exercise.

METHODS

We evaluated exDSS using various exercise and meal scenarios recorded from a large, free-living study of aerobic and resistance exercise. The model inputs were heart rate, insulin, and meal data. Glucose responses were simulated during and after 30-minute exercise sessions (676 aerobic, 631 resistance) from 247 participants. Glucose outcomes were compared when participants followed exDSS recommendations, clinical guidelines, or did not modify behavior (no intervention).

RESULTS

exDSS significantly improved mean time in range for aerobic (80.2% to 92.3%, P < .0001) and resistance (72.3% to 87.3%, P < .0001) exercises compared with no intervention, and versus clinical guidelines (aerobic: 82.2%, P < .0001; resistance: 80.3%, P < .0001). exDSS reduced time spent in low glucose for both exercise types compared with no intervention (aerobic: 15.1% to 5.1%, P < .0001; resistance: 18.2% to 6.6%, P < .0001) and was comparable with following clinical guidelines (aerobic: 4.5%, resistance: 8.1%, P = N.S.).

CONCLUSIONS

The exDSS tool significantly improved glucose outcomes during and after exercise versus following clinical guidelines and no intervention providing motivation for clinical evaluation of the exDSS system.

Artificial intelligence in critically ill diabetic patients: current status and future prospects

Recent years have witnessed increasing numbers of artificial intelligence (AI) based applications and devices being tested and approved for medical care. Diabetes is arguably the most common chronic disorder worldwide and AI is now being used for making an early diagnosis, to predict and diagnose early complications, increase adherence to therapy, and even motivate patients to manage diabetes and maintain glycemic control. However, these AI applications have largely been tested in non-critically ill patients and aid in managing chronic problems. Intensive care units (ICUs) have a dynamic environment generating huge data, which AI can extract and organize simultaneously, thus analysing many variables for diagnostic and/or therapeutic purposes in order to predict outcomes of interest. Even non-diabetic ICU patients are at risk of developing hypo or hyperglycemia, complicating their ICU course and affecting outcomes. In addition, to maintain glycemic control frequent blood sampling and insulin dose adjustments are required, increasing nursing workload and chances of error. AI has the potential to improve glycemic control while reducing the nursing workload and errors. Continuous glucose monitoring (CGM) devices, which are Food and Drug Administration (FDA) approved for use in non-critically ill patients, are now being recommended for use in specific ICU populations with increased accuracy. AI based devices including artificial pancreas and CGM regulated insulin infusion system have shown promise as comprehensive glycemic control solutions in critically ill patients. Even though many of these AI applications have shown potential, these devices need to be tested in larger number of ICU patients, have wider availability, show favorable cost-benefit ratio and be amenable for easy integration into the existing healthcare systems, before they become acceptable to ICU physicians for routine use.

A Model-Based Insulin Dose Optimization Algorithm for People With Type 1 Diabetes on Multiple Daily Injections Therapy

OBJECTIVE

Multiple daily injections (MDI) therapy is the most common treatment for type 1 diabetes (T1D) including basal insulin doses to keep glucose levels constant during fasting conditions and bolus insulin doses with meals. Optimal insulin dosing is critical to achieving satisfactory glycemia but is challenging due to inter- and intra-individual variability. Here, we present a novel model-based iterative algorithm that optimizes insulin doses using previous-day glucose, insulin, and meal data.

METHODS

Our algorithm employs a maximum-a-posteriori method to estimate parameters of a model describing the effects of changes in basal-bolus insulin doses. Then, parameter estimates, their confidence intervals, and the goodness of fit, are combined to generate new recommendations. We assessed our algorithm in three ways. First, a clinical data set of 150 days (15 participants) were used to evaluate the proposed model and the estimation method. Second, 60-day simulations were performed to demonstrate the efficacy of the algorithm. Third, a sample 6-day clinical experiment is presented and discussed.

RESULTS

The model fitted the clinical data well with a root-mean-square-error of 1.75 mmol/L. Simulation results showed an improvement in the time in target (3.9-10 mmol/L) from 64% to 77% and a decrease in the time in hypoglycemia (< 3.9 mmol/L) from 8.1% to 3.8%. The clinical experiment demonstrated the feasibility of the algorithm.

CONCLUSION

Our algorithm has the potential to improve glycemic control in people with T1D using MDI.

SIGNIFICANCE

This work is a step forward towards a decision support system that improves their quality of life.

An Insulin Bolus Advisor for Type 1 Diabetes Using Deep Reinforcement Learning

(1) Background: People living with type 1 diabetes (T1D) require self-management to maintain blood glucose (BG) levels in a therapeutic range through the delivery of exogenous insulin. However, due to the various variability, uncertainty and complex glucose dynamics, optimizing the doses of insulin delivery to minimize the risk of hyperglycemia and hypoglycemia is still an open problem. (2) Methods: In this work, we propose a novel insulin bolus advisor which uses deep reinforcement learning (DRL) and continuous glucose monitoring to optimize insulin dosing at mealtime. In particular, an actor-critic model based on deep deterministic policy gradient is designed to compute mealtime insulin doses. The proposed system architecture uses a two-step learning framework, in which a population model is first obtained and then personalized by subject-specific data. Prioritized memory replay is adopted to accelerate the training process in clinical practice. To validate the algorithm, we employ a customized version of the FDA-accepted UVA/Padova T1D simulator to perform in silico trials on 10 adult subjects and 10 adolescent subjects. (3) Results: Compared to a standard bolus calculator as the baseline, the DRL insulin bolus advisor significantly improved the average percentage time in target range (70-180 mg/dL) from 74.1%±8.4% to 80.9%±6.9% (p<0.01) and 54.9%±12.4% to 61.6%±14.1% (p<0.01) in the the adult and adolescent cohorts, respectively, while reducing hypoglycemia. (4) Conclusions: The proposed algorithm has the potential to improve mealtime bolus insulin delivery in people with T1D and is a feasible candidate for future clinical validation.

Advanced Diabetes Management Using Artificial Intelligence and Continuous Glucose Monitoring Sensors

Wearable continuous glucose monitoring (CGM) sensors are revolutionizing the treatment of type 1 diabetes (T1D). These sensors provide in real-time, every 1-5 min, the current blood glucose concentration and its rate-of-change, two key pieces of information for improving the determination of exogenous insulin administration and the prediction of forthcoming adverse events, such as hypo-/hyper-glycemia. The current research in diabetes technology is putting considerable effort into developing decision support systems for patient use, which automatically analyze the patient's data collected by CGM sensors and other portable devices, as well as providing personalized recommendations about therapy adjustments to patients. Due to the large amount of data collected by patients with T1D and their variety, artificial intelligence (AI) techniques are increasingly being adopted in these decision support systems. In this paper, we review the state-of-the-art methodologies using AI and CGM sensors for decision support in advanced T1D management, including techniques for personalized insulin bolus calculation, adaptive tuning of bolus calculator parameters and glucose prediction.

An artificial intelligence decision support system for the management of type 1 diabetes

Type 1 diabetes (T1D) is characterized by pancreatic beta cell dysfunction and insulin depletion. Over 40% of people with T1D manage their glucose through multiple injections of long-acting basal and short-acting bolus insulin, so-called multiple daily injections (MDI)^1,2. Errors in dosing can lead to life-threatening hypoglycaemia events (<70 mg dl^-1) and hyperglycaemia (>180 mg dl^-1), increasing the risk of retinopathy, neuropathy, and nephropathy. Machine learning (artificial intelligence) approaches are being harnessed to incorporate decision support into many medical specialties. Here, we report an algorithm that provides weekly insulin dosage recommendations to adults with T1D using MDI therapy. We employ a unique virtual platform³ to generate over 50,000 glucose observations to train a k-nearest neighbours⁴ decision support system (KNN-DSS) to identify causes of hyperglycaemia or hypoglycaemia and determine necessary insulin adjustments from a set of 12 potential recommendations. The KNN-DSS algorithm achieves an overall agreement with board-certified endocrinologists of 67.9% when validated on real-world human data, and delivers safe recommendations, per endocrinologist review. A comparison of inter-physician-recommended adjustments to insulin pump therapy indicates full agreement of 41.2% among endocrinologists, which is consistent with previous measures of inter-physician agreement (41-45%)⁵. In silico^3,6 benchmarking using a platform accepted by the United States Food and Drug Administration for evaluation of artificial pancreas technologies indicates substantial improvement in glycaemic outcomes after 12 weeks of KNN-DSS use. Our data indicate that the KNN-DSS allows for early identification of dangerous insulin regimens and may be used to improve glycaemic outcomes and prevent life-threatening complications in people with T1D.

Artificial Intelligence in Decision Support Systems for Type 1 Diabetes

Type 1 diabetes (T1D) is a chronic health condition resulting from pancreatic beta cell dysfunction and insulin depletion. While automated insulin delivery systems are now available, many people choose to manage insulin delivery manually through insulin pumps or through multiple daily injections. Frequent insulin titrations are needed to adequately manage glucose, however, provider adjustments are typically made every several months. Recent automated decision support systems incorporate artificial intelligence algorithms to deliver personalized recommendations regarding insulin doses and daily behaviors. This paper presents a comprehensive review of computational and artificial intelligence-based decision support systems to manage T1D. Articles were obtained from PubMed, IEEE Xplore, and ScienceDirect databases. No time period restrictions were imposed on the search. After removing off-topic articles and duplicates, 562 articles were left to review. Of those articles, we identified 61 articles for comprehensive review based on algorithm evaluation using real-world human data, in silico trials, or clinical studies. We grouped decision support systems into general categories of (1) those which recommend adjustments to insulin and (2) those which predict and help avoid hypoglycemia. We review the artificial intelligence methods used for each type of decision support system, and discuss the performance and potential applications of these systems.

Reinforcement Learning-Based Adaptive Insulin Advisor for Individuals with Type 1 Diabetes Patients under Multiple Daily Injections Therapy

The existing adaptive basal-bolus advisor (ABBA) was further developed to benefit patients under insulin therapy with multiple daily injections (MDI). Three different in silico experiments were conducted with the DMMS.R simulator to validate the approach of combined use of self-monitoring of blood glucose (SMBG) and insulin injection devices, e.g. insulin pen, as are used by the majority of type 1 diabetes patients under insulin therapy. The proposed approach outperforms the conventional method, as it increases the time spent within the target range and simultaneously reduces the risks of hyperglycaemic and hypoglycaemic events.

Intelligent Data-Driven Model for Diabetes Diurnal Patterns Analysis

In type 1 diabetes, diurnal activity routines are influential factors in insulin dose calculations. Bolus advisors have been developed to more accurately suggest doses of meal-related insulin based on carbohydrate intake, according to pre-set insulin to carbohydrate levels and insulin sensitivity factors. These parameters can be varied according to the time of day and their optimal setting relies on identifying the daily time periods of routines accurately. The main issues with reporting and adjustments of daily activity routines are the reliance on self-reporting which is prone to inaccuracy and within bolus calculators, the keeping of default settings for daily time periods, such as within insulin pumps, glucose meters, and mobile applications. Moreover, daily routines are subject to change over periods of time which could go unnoticed. Hence, forgetting to change the daily time periods in the bolus calculator could contribute to sub-optimal self-management. In this paper, these issues are addressed by proposing a data-driven model for identification of diabetes diurnal patterns based on self-monitoring data. The model uses time-series clustering to achieve a meaningful separation of the patterns which is then used to identify the daily time periods and to advise of any time changes required. Further improvements in bolus advisor settings are proposed to include week/weekend or even modifiable daily time settings. The proposed model provides a quick, granular, more accurate, and personalized daily time setting profile while providing a more contextual perspective to glycemic pattern identification to both patients and clinicians.

Control-Oriented Model With Intra-Patient Variations for an Artificial Pancreas

In this work, a low-order model designed for glucose regulation in Type 1 Diabetes Mellitus (T1DM) is obtained from the UVA/Padova metabolic simulator. It captures not only the nonlinear behavior of the glucose-insulin system, but also intra-patient variations related to daily insulin sensitivity ( S_I) changes. To overcome the large inter-subject variability, the model can also be personalized based on a priori patient information. The structure is amenable for linear parameter varying (LPV) controller design, and represents the dynamics from the subcutaneous insulin input to the subcutaneous glucose output. The efficacy of this model is evaluated in comparison with a previous control-oriented model which in turn is an improvement of previous models. Both models are compared in terms of their open- and closed-loop differences with respect to the UVA/Padova model. The proposed model outperforms previous T1DM control-oriented models, which could potentially lead to more robust and reliable controllers for glycemia regulation.

A Modular Safety System for an Insulin Dose Recommender: A Feasibility Study

BACKGROUND

Delivering insulin in type 1 diabetes is a challenging, and potentially risky, activity; hence the importance of including safety measures as part of any insulin dosing or recommender system. This work presents and clinically evaluates a modular safety system that is part of an intelligent insulin dose recommender platform developed within the EU-funded PEPPER project.

METHODS

The proposed safety system is composed of four modules which use a novel glucose forecasting algorithm. These modules are predictive glucose alerts and alarms; a predictive low-glucose basal insulin suspension module; an advanced rescue carbohydrate recommender for resolving hypoglycemia; and a personalized safety constraint applied to insulin recommendations. The technical feasibility of the proposed safety system was evaluated in a pilot study including eight adult subjects with type 1 diabetes on multiple daily injections over a duration of six weeks. Glycemic control and safety system functioning were compared between the two-weeks run-in period and the end point at eight weeks. A standard insulin bolus calculator was employed to recommend insulin doses.

RESULTS

Overall, glycemic control improved over the evaluated period. In particular, percentage time in the hypoglycemia range (<3.0 mmol/l) significantly decreased from 0.82% (0.05-4.79) at run-in to 0.33% (0.00-0.93) at endpoint (P = .02). This was associated with a significant increase in percentage time in target range (3.9-10.0 mmol/l) from 52.8% (38.3-61.5) to 61.3% (47.5-71.7) (P = .03). There was also a reduction in number of carbohydrate recommendations.

CONCLUSION

A safety system for an insulin dose recommender has been proven to be a viable solution to reduce the number of adverse events associated to glucose control in type 1 diabetes.

Application of Artificial Intelligence Techniques for the Estimation of Basal Insulin in Patients with Type I Diabetes

Artificial intelligence techniques have been positioned in the resolution of problems in various areas of healthcare. Clinical decision support systems developed from this technology have optimized the healthcare of patients with chronic diseases through mobile applications. In this study, several models based on this methodology have been developed to calculate the basal insulin dose in patients with type I diabetes using subcutaneous insulin infusion pumps. Methods. A pilot experimental study was performed with data from 56 patients with type 1 diabetes who used insulin infusion pumps and underwent continuous glucose monitoring. Several models based on artificial intelligence techniques were developed to analyze glycemic patterns based on continuous glucose monitoring and clinical variables in order to estimate the basal insulin dose. We used neural networks (NNs), Bayesian networks (BNs), support vector machines (SVMs), and random forests (RF). We then evaluated the agreement between predicted and actual values using several statistical error measurements: mean absolute error (MAE), mean square error (MSE), root-mean-square error (RMSE), Pearson's correlation coefficient (R), and determination coefficient (R ²). Results. Twenty-four different models were obtained, one for each hour of the day, with each chosen technique. Correlation coefficients obtained with RF, SVMs, NNs, and BNs were 0.9999, 0.9921, 0.0303, and 0.7754, respectively. The error increased between 06:00 and 07:00 and between 13:00 and 17:00. Conclusions. The performance of the RF technique was excellent and got very close to the actual values. Intelligence techniques could be used to predict basal insulin dose. However, it is necessary to explore the validity of the results and select the target population. Models that allow for more accurate levels of prediction should be further explored.

Continuous Glucose Monitors and Activity Trackers to Inform Insulin Dosing in Type 1 Diabetes: The University of Virginia Contribution

Objective: Suboptimal insulin dosing in type 1 diabetes (T1D) is frequently associated with time-varying insulin requirements driven by various psycho-behavioral and physiological factors influencing insulin sensitivity (IS). Among these, physical activity has been widely recognized as a trigger of altered IS both during and following the exercise effort, but limited indication is available for the management of structured and (even more) unstructured activity in T1D. In this work, we present two methods to inform insulin dosing with biosignals from wearable sensors to improve glycemic control in individuals with T1D. Research Design and Methods: Continuous glucose monitors (CGM) and activity trackers are leveraged by the methods. The first method uses CGM records to estimate IS in real time and adjust the insulin dose according to a person’s insulin needs; the second method uses step count data to inform the bolus calculation with the residual glucose-lowering effects of recently performed (structured or unstructured) physical activity. The methods were tested in silico within the University of Virginia/Padova T1D Simulator. A standard bolus calculator and the proposed “smart” systems were deployed in the control of one meal in presence of increased/decreased IS (Study 1) and following a 1-hour exercise bout (Study 2). Postprandial glycemic control was assessed in terms of time spent in different glycemic ranges and low/high blood glucose indices (LBGI/HBGI), and compared between the dosing strategies. Results: In Study 1, the CGM-informed system allowed to reduce exposure to hypoglycemia in presence of increased IS (percent time < 70 mg/dL: 6.1% versus 9.9%; LBGI: 1.9 versus 3.2) and exposure to hyperglycemia in presence of decreased IS (percent time > 180 mg/dL: 14.6% versus 18.3%; HBGI: 3.0 versus 3.9), tending toward optimal control. In Study 2, the step count-informed system allowed to reduce hypoglycemia (percent time < 70 mg/dL: 3.9% versus 13.4%; LBGI: 1.7 versus 3.2) at the cost of a minor increase in exposure to hyperglycemia (percent time > 180 mg/dL: 11.9% versus 7.5%; HBGI: 2.4 versus 1.5). Conclusions: We presented and validated in silico two methods for the smart dosing of prandial insulin in T1D. If seen within an ensemble, the two algorithms provide alternatives to individuals with T1D for improving insulin dosing accommodating a large variety of treatment options. Future work will be devoted to test the safety and efficacy of the methods in free-living conditions.

Detection of Insulin Pump Malfunctioning to Improve Safety in Artificial Pancreas Using Unsupervised Algorithms

BACKGROUND

Recent development of automated closed-loop (CL) insulin delivery systems, the so-called artificial pancreas (AP), improved the quality of type 1 diabetes (T1D) therapy. As new technologies emerge, patients put increasing trust in their therapeutic devices; therefore, it becomes increasingly important to detect malfunctioning affecting such devices. In this work, we explore a new paradigm to detect insulin pump faults (IPFs) that use unsupervised anomaly detection.

METHODS

We generated CL data corrupted with IPFs using the latest version of the T1D Padova/UVA simulator. From the data, we extracted several features capable to describe the patient dynamics and making more apparent suspicious data portions. Then, a feature selection is performed to determine the optimal feature set. Finally, the performance of several popular unsupervised anomaly detection algorithms is analyzed and compared on the identified optimal feature set.

RESULTS

Using the identified optimal configuration, the best performance is obtained by the Histogram-Based Outlier Score (HBOS) algorithm, which detected 87% of the IPF with only 0.08 false positives per day on average. Isolation forest is the best algorithm that offers more conservative performances, detection of 85% of the faults but only 0.06 false positives per day on average.

CONCLUSION

Unsupervised anomaly detection algorithms can be used effectively to detect IPFs and improve the safety of the AP. Future studies will be dedicated to test the presented method inside dedicated clinical trials.

Predicting Quality of Overnight Glycaemic Control in Type 1 Diabetes Using Binary Classifiers

In type 1 diabetes management, maintaining nocturnal blood glucose within target range can be challenging. Although semi-automatic systems to modulate insulin pump delivery, such as low-glucose insulin suspension and the artificial pancreas, are starting to become a reality, their elevated cost and performance below user expectations is hindering their adoption. Hence, a decision support system that helps people with type 1 diabetes, on multiple daily injections or insulin pump therapy, to avoid undesirable overnight blood glucose fluctuations (hyper- or hypoglycaemic) is an attractive alternative. In this paper, we introduce a novel data-driven approach to predict the quality of overnight glycaemic control in people with type 1 diabetes by analyzing commonly gathered data during the day-time period (continuous glucose monitoring data, meal intake and insulin boluses). The proposed approach is able to predict whether overnight blood glucose concentrations are going to remain within or outside the target range, and therefore allows the user to take the appropriate preventive action (snack or change in basal insulin). For this purpose, a number of popular established machine learning algorithms for binary classification were evaluated and compared on a publicly available clinical dataset (i.e., OhioT1DM). Although there is no clearly superior classification algorithm, this study indicates that, by using commonly gathered data in type 1 diabetes management, it is possible to predict the quality of overnight glycaemic control with reasonable accuracy (AUC-ROC = 0.7).

Classification of Postprandial Glycemic Status with Application to Insulin Dosing in Type 1 Diabetes-An In Silico Proof-of-Concept

In the daily management of type 1 diabetes (T1D), determining the correct insulin dose to be injected at meal-time is fundamental to achieve optimal glycemic control. Wearable sensors, such as continuous glucose monitoring (CGM) devices, are instrumental to achieve this purpose. In this paper, we show how CGM data, together with commonly recorded inputs (carbohydrate intake and bolus insulin), can be used to develop an algorithm that allows classifying, at meal-time, the post-prandial glycemic status (i.e., blood glucose concentration being too low, too high, or within target range). Such an outcome can then be used to improve the efficacy of insulin therapy by reducing or increasing the corresponding meal bolus dose. A state-of-the-art T1D simulation environment, including intraday variability and a behavioral model, was used to generate a rich in silico dataset corresponding to 100 subjects over a two-month scenario. Then, an extreme gradient-boosted tree (XGB) algorithm was employed to classify the post-prandial glycemic status. Finally, we demonstrate how the XGB algorithm outcome can be exploited to improve glycemic control in T1D through real-time adjustment of the meal insulin bolus. The proposed XGB algorithm obtained good accuracy at classifying post-prandial glycemic status (AUROC = 0.84 [0.78, 0.87]). Consequently, when used to adjust, in real-time, meal insulin boluses obtained with a bolus calculator, the proposed approach improves glycemic control when compared to the baseline bolus calculator. In particular, percentage time in target [70, 180] mg/dL was improved from 61.98 (±13.89) to 67.00 (±11.54; p < 0.01) without increasing hypoglycemia.

A Dual Mode Adaptive Basal-Bolus Advisor Based on Reinforcement Learning

Self-monitoring of blood glucose (SMBG) and continuous glucose monitoring (CGM) are commonly used by type 1 diabetes (T1D) patients to measure glucose concentrations. The proposed adaptive basal-bolus algorithm (ABBA) supports inputs from either SMBG or CGM devices to provide personalised suggestions for the daily basal rate and prandial insulin doses on the basis of the patients' glucose level on the previous day. The ABBA is based on reinforcement learning, a type of artificial intelligence, and was validated in silico with an FDA-accepted population of 100 adults under different realistic scenarios lasting three simulated months. The scenarios involve three main meals and one bedtime snack per day, along with different variabilities and uncertainties for insulin sensitivity, mealtime, carbohydrate amount, and glucose measurement time. The results indicate that the proposed approach achieves comparable performance with CGM or SMBG as input signals, without influencing the total daily insulin dose. The results are a promising indication that AI algorithmic approaches can provide personalised adaptive insulin optimization and achieve glucose control-independent of the type of glucose monitoring technology.

Transforming Diabetes Care Through Artificial Intelligence: The Future Is Here

An estimated 425 million people globally have diabetes, accounting for 12% of the world's health expenditures, and yet 1 in 2 persons remain undiagnosed and untreated. Applications of artificial intelligence (AI) and cognitive computing offer promise in diabetes care. The purpose of this article is to better understand what AI advances may be relevant today to persons with diabetes (PWDs), their clinicians, family, and caregivers. The authors conducted a predefined, online PubMed search of publicly available sources of information from 2009 onward using the search terms "diabetes" and "artificial intelligence." The study included clinically-relevant, high-impact articles, and excluded articles whose purpose was technical in nature. A total of 450 published diabetes and AI articles met the inclusion criteria. The studies represent a diverse and complex set of innovative approaches that aim to transform diabetes care in 4 main areas: automated retinal screening, clinical decision support, predictive population risk stratification, and patient self-management tools. Many of these new AI-powered retinal imaging systems, predictive modeling programs, glucose sensors, insulin pumps, smartphone applications, and other decision-support aids are on the market today with more on the way. AI applications have the potential to transform diabetes care and help millions of PWDs to achieve better blood glucose control, reduce hypoglycemic episodes, and reduce diabetes comorbidities and complications. AI applications offer greater accuracy, efficiency, ease of use, and satisfaction for PWDs, their clinicians, family, and caregivers.

Continuous Glucose Monitoring and Insulin Informed Advisory System with Automated Titration and Dosing of Insulin Reduces Glucose Variability in Type 1 Diabetes Mellitus

BACKGROUND

Glucose variability (GV) remains a key limiting factor in the success of diabetes management. While new technologies, for example, accurate continuous glucose monitoring (CGM) and connected insulin delivery devices, are now available, current treatment standards fail to leverage the wealth of information generated. Expert systems, from automated insulin delivery to advisory systems, are a key missing element to richer, more personalized, glucose management in diabetes.

METHODS

Twenty four subjects with type 1 diabetes mellitus (T1DM), 15 women, 37 ± 11 years of age, hemoglobin A1c 7.2% ± 1%, total daily insulin (TDI) 46.7 ± 22.3 U, using either an insulin pump or multiple daily injections with carbohydrate counting, completed two randomized crossover 48-h visits at the University of Virginia, wearing Dexcom G4 CGM, and using either usual care or the UVA decision support system (DSS). DSS consisted of a combination of automated insulin titration, bolus calculation, and CHO treatment advice. During each admission, participants were exposed to a variety of meal sizes and contents and two 45-min bouts of exercise. GV and glucose control were assessed using CGM.

RESULTS

The use of DSS significantly reduced GV (coefficient of variation: 0.36 ± 08. vs. 0.33 ± 0.06, P = 0.045) while maintaining glycemic control (average CGM: 155.2 ± 27.1 mg/dL vs. 155.2 ± 23.2 mg/dL), by reducing hypoglycemia exposure (%<70 mg/dL: 3.8% ± 4.6% vs. 1.8% ± 2%, P = 0.018), with nonsignificant trends toward reduction of significant hyperglycemia overnight (%>250 mg/dL: 5.3% ± 9.5% vs. 1.9% ± 4.6%) and at mealtime (11.3% ± 14.8% vs. 5.8% ± 9.1%).

CONCLUSIONS

A CGM/insulin informed advisory system proved to be safe and feasible in a cohort of 24 T1DM subjects. Use of the system may result in reduced GV and improved protection against hypoglycemia.

Artificial Intelligence for Diabetes Management and Decision Support: Literature Review

Background

Artificial intelligence methods in combination with the latest technologies, including medical devices, mobile computing, and sensor technologies, have the potential to enable the creation and delivery of better management services to deal with chronic diseases. One of the most lethal and prevalent chronic diseases is diabetes mellitus, which is characterized by dysfunction of glucose homeostasis.

Objective

The objective of this paper is to review recent efforts to use artificial intelligence techniques to assist in the management of diabetes, along with the associated challenges.

Methods

A review of the literature was conducted using PubMed and related bibliographic resources. Analyses of the literature from 2010 to 2018 yielded 1849 pertinent articles, of which we selected 141 for detailed review.

Results

We propose a functional taxonomy for diabetes management and artificial intelligence. Additionally, a detailed analysis of each subject category was performed using related key outcomes. This approach revealed that the experiments and studies reviewed yielded encouraging results.

Conclusions

We obtained evidence of an acceleration of research activity aimed at developing artificial intelligence-powered tools for prediction and prevention of complications associated with diabetes. Our results indicate that artificial intelligence methods are being progressively established as suitable for use in clinical daily practice, as well as for the self-management of diabetes. Consequently, these methods provide powerful tools for improving patients’ quality of life.

Delivering precision antimicrobial therapy through closed-loop control systems

Sub-optimal exposure to antimicrobial therapy is associated with poor patient outcomes and the development of antimicrobial resistance. Mechanisms for optimizing the concentration of a drug within the individual patient are under development. However, several barriers remain in realizing true individualization of therapy. These include problems with plasma drug sampling, availability of appropriate assays, and current mechanisms for dose adjustment. Biosensor technology offers a means of providing real-time monitoring of antimicrobials in a minimally invasive fashion. We report the potential for using microneedle biosensor technology as part of closed-loop control systems for the optimization of antimicrobial therapy in individual patients.

Personalized Adaptive CBR Bolus Recommender System for Type 1 Diabetes

Type 1 diabetes mellitus (T1DM) is a chronic disease. Those who have it must administer themselves with insulin to control their blood glucose level. It is difficult to estimate the correct insulin dosage due to the complex glucose metabolism, which can lead to less than optimal blood glucose levels. This paper presents PepperRec, a case-based reasoning (CBR) bolus insulin recommender system capable of dealing with an unrestricted number of situations in which T1DM persons can find themselves. PepperRec considers several factors that affect glucose metabolism, such as data about the physical activity of the user, and can also cope with missing values for these factors. Based on CBR methodology, PepperRec uses new methods to adapt past recommendations to the current state of the user, and retains updated historical patient information to deal with slow and gradual changes in the patient over time (concept drift). The proposed approach is tested using the UVA/PADOVA simulator with 33 virtual subjects and compared with other methods in the literature, and with the default insulin therapy of the simulator. The achieved results demonstrate that PepperRec increases the amount of time the users are in their target glycaemic range, reduces the time spent below it, while maintaining, or even reducing, the time spent above it.

Automatic Adaptation of Basal Insulin Using Sensor-Augmented Pump Therapy

BACKGROUND

People with insulin-dependent diabetes rely on an intensified insulin regimen. Despite several guidelines, they are usually impractical and fall short in achieving optimal glycemic outcomes. In this work, a novel technique for automatic adaptation of the basal insulin profile of people with diabetes on sensor-augmented pump therapy is presented.

METHODS

The presented technique is based on a run-to-run control law that overcomes some of the limitations of previously proposed methods. To prove its validity, an in silico validation was performed. Finally, the artificial intelligence technique of case-based reasoning is proposed as a potential solution to deal with variability in basal insulin requirements.

RESULTS

Over a period of 4 months, the proposed run-to-run control law successfully adapts the basal insulin profile of a virtual population (10 adults, 10 adolescents, and 10 children). In particular, average percentage time in target [70, 180] mg/dl was significantly improved over the evaluated period (first week versus last week): 70.9 ± 11.8 versus 91.1 ± 4.4 (adults), 46.5 ± 11.9 versus 80.1 ± 10.9 (adolescents), 49.4 ± 12.9 versus 73.7 ± 4.1 (children). Average percentage time in hypoglycemia (<70 mg/dl) was also significantly reduced: 9.7 ± 6.6 versus 0.9 ± 1.2 (adults), 10.5 ± 8.3 versus 0.83 ± 1.0 (adolescents), 10.9 ± 6.1 versus 3.2 ± 3.5 (children). When compared against an existing technique over the whole evaluated period, the presented approach achieved superior results on percentage of time in hypoglycemia: 3.9 ± 2.6 versus 2.6 ± 2.2 (adults), 2.9 ± 1.9 versus 2.0 ± 1.5 (adolescents), 4.6 ± 2.8 versus 3.5 ± 2.0 (children), without increasing the percentage time in hyperglycemia.

CONCLUSION

The present study shows the potential of a novel technique to effectively adjust the basal insulin profile of a type 1 diabetes population on sensor-augmented insulin pump therapy.

Insulin Sensitivity Index-Based Optimization of Insulin to Carbohydrate Ratio: In Silico Study Shows Efficacious Protection Against Hypoglycemic Events Caused by Suboptimal Therapy

BACKGROUND AND AIM

The insulin to carbohydrate ratio (CR) is a parameter used by patients with type 1 diabetes (T1D) to calculate the premeal insulin bolus. Usually, it is estimated by the physician based on patient diary, but modern diabetes technologies, such as subcutaneous glucose sensing (continuous glucose monitoring, CGM) and insulin delivery (continuous subcutaneous insulin infusion, CSII) systems, can provide important information for its optimization. In this study, a method for CR optimization based on CGM and CSII data is presented and its efficacy and robustness tested in several in silico scenarios, with the primary aim of increasing protection against hypoglycemia.

METHODS

The method is based on a validated index of insulin sensitivity calculated from sensor and pump data (S_I^SP), area under CGM and CSII curves. The efficacy and robustness of the method are tested in silico using the University of Virginia/Padova T1D simulator, in several suboptimal therapy scenarios: with nominal CR variation, over/underestimation of meal size or suboptimal basal insulin infusion. Simulated CGM and CSII data were used to calculate the optimal CR. The same scenarios were then repeated using the estimated CR and glycemic control was compared.

RESULTS

The optimized CR was efficacious in protecting against hypoglycemic events caused by suboptimal therapy. The method was also robust to possible error in carbohydrate count and suboptimal basal insulin infusion.

CONCLUSIONS

A novel method for CR optimization in T1D, implementable in daily life using CGM and CSII data, is proposed. The method can be used both in open- and closed-loop insulin therapy.

Machine Learning and Data Mining Methods in Diabetes Research

The remarkable advances in biotechnology and health sciences have led to a significant production of data, such as high throughput genetic data and clinical information, generated from large Electronic Health Records (EHRs). To this end, application of machine learning and data mining methods in biosciences is presently, more than ever before, vital and indispensable in efforts to transform intelligently all available information into valuable knowledge. Diabetes mellitus (DM) is defined as a group of metabolic disorders exerting significant pressure on human health worldwide. Extensive research in all aspects of diabetes (diagnosis, etiopathophysiology, therapy, etc.) has led to the generation of huge amounts of data. The aim of the present study is to conduct a systematic review of the applications of machine learning, data mining techniques and tools in the field of diabetes research with respect to a) Prediction and Diagnosis, b) Diabetic Complications, c) Genetic Background and Environment, and e) Health Care and Management with the first category appearing to be the most popular. A wide range of machine learning algorithms were employed. In general, 85% of those used were characterized by supervised learning approaches and 15% by unsupervised ones, and more specifically, association rules. Support vector machines (SVM) arise as the most successful and widely used algorithm. Concerning the type of data, clinical datasets were mainly used. The title applications in the selected articles project the usefulness of extracting valuable knowledge leading to new hypotheses targeting deeper understanding and further investigation in DM.

Case-Based Reasoning for Insulin Bolus Advice

BACKGROUND

Insulin bolus calculators assist people with Type 1 diabetes (T1D) to calculate the amount of insulin required for meals to achieve optimal glucose levels but lack adaptability and personalization. We have proposed enhancing bolus calculators by the means of case-based reasoning (CBR), an established problem-solving methodology, by individualizing and optimizing insulin therapy for various meal situations. CBR learns from experiences of past similar meals, which are described in cases through a set of parameters (eg, time of meal, alcohol, exercise). This work discusses the selection, representation and effect of case parameters used for a CBR-based Advanced Bolus Calculator for Diabetes (ABC4D).

METHODS

We analyzed the usage and effect of selected parameters during a pilot study (n = 10), where participants used ABC4D for 6 weeks. Retrospectively, we evaluated the effect of glucose rate of change before the meal on the glycemic excursion. Feedback from study participants about the choice of parameters was obtained through a nonvalidated questionnaire.

RESULTS

Exercise and alcohol were the most frequently used parameters, which was congruent with the feedback from study participants, who found these parameters most useful. Furthermore, cases including either exercise or alcohol as parameter showed a trend in reduction of insulin at the end of the study. A significant difference ( P < .01) was found in glycemic outcomes for meals where glucose rate of change was rising compared to stable rate of change.

CONCLUSIONS

Results from the 6-week study indicate the potential benefit of including parameters exercise, alcohol and glucose-rate of change for insulin dosing decision support.

An Advanced Bolus Calculator for Type 1 Diabetes: System Architecture and Usability Results

This paper presents the architecture and initial usability results of an advanced insulin bolus calculator for diabetes (ABC4D), which provides personalized insulin recommendations for people with diabetes by differentiating between various diabetes scenarios and automatically adjusting its parameters over time. The proposed platform comprises two main components: a smartphone-based patient platform allowing manual input of glucose and variables affecting blood glucose levels (e.g., meal carbohydrate content and exercise) and providing real-time insulin bolus recommendations; and a clinical revision platform to supervise the automatic adaptations of the bolus calculator parameters. The system implements a previously in silico validated bolus calculator algorithm based on case-based reasoning, which uses information from similar past events (i.e., cases) to suggest improved personalized insulin bolus recommendations and automatically learns from new events. Usability of ABC4D was assessed by analyzing the system usage at the end of a six-week pilot study (n = 10). Further feedback on the use of ABC4D has been obtained from each participant at the end of the study from a usability questionnaire. On average, each participant requested 115 ± 21 insulin recommendations, of which 103 ± 28 (90%) were accepted. The clinical revision software proposed a total of 754 case revisions, where 723 (96%) adaptations were approved by a clinical expert and updated in the patient platform.