Incorporating an Exercise Detection, Grading, and Hormone Dosing Algorithm Into the Artificial Pancreas Using Accelerometry and Heart Rate

Glucose-responsive insulin delivery for type 1 diabetes: The artificial pancreas story

Insulin replacement therapy is integral to the management of type 1 diabetes, which is characterised by absolute insulin deficiency. Optimal glycaemic control, as assessed by glycated haemoglobin, and avoidance of hyper- and hypoglycaemic excursions have been shown to prevent diabetes-related complications. Insulin pump use has increased considerably over the past decade with beneficial effects on glycaemic control, quality of life and treatment satisfaction. The advent and progress of ambulatory glucose sensor technology has enabled continuous glucose monitoring based on real-time glucose levels to be integrated with insulin therapy. Low glucose and predictive low glucose suspend systems are currently used in clinical practice to mitigate against hypoglycaemia, and provide the first step towards feedback glucose control. The more advanced technology approach, an artificial pancreas or a closed-loop system, gradually increases and decreases insulin delivery in a glucose-responsive fashion to mitigate against hyper- and hypoglycaemia. Randomised outpatient clinical trials over the past 5 years have demonstrated the feasibility, safety and efficacy of the approach, and the recent FDA approval of the first single hormone closed-loop system establishes a new standard of care for people with type 1 diabetes.

Multivariable Adaptive Artificial Pancreas System in Type 1 Diabetes

PURPOSE OF REVIEW

The review summarizes the current state of the artificial pancreas (AP) systems and introduces various new modules that should be included in future AP systems.

RECENT FINDINGS

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

Design of a dual-hormone model predictive control for artificial pancreas with exercise model

The Artificial Pancreas (AP) is a new technology for helping people with type 1 diabetes to better control their glucose levels through automated delivery of insulin and optionally glucagon in response to sensed glucose levels. In a dual hormone AP, insulin and glucagon are delivered automatically to the body based on glucose sensor measurements using a control algorithm that calculates the amount of hormones to be infused. A dual-hormone MPC may deliver insulin continuously; however, it must avoid continuous delivery of glucagon because nausea can occur from too much glucagon. In this paper, we propose a novel dual-hormone (DH) switching model predictive control and compare it with a single-hormone (SH) MPC. We extended both MPCs by integrating an exercise model and compared performance with and without the exercise model included. Results were obtained on a virtual patient population undergoing a simulated exercise event using a mathematical glucoregulatory model that includes exercise. Time spent in hypoglycemia is significantly less with the DH-MPC than the SH-MPC (p=0.0022). Additionally, including the exercise model in the DH-MPC can help prevent hypoglycemia (p <; 0.001).

An original piecewise model for computing energy expenditure from accelerometer and heart rate signals

OBJECTIVE

Activity energy expenditure (EE) plays an important role in healthcare, therefore, accurate EE measures are required. Currently available reference EE acquisition methods, such as doubly labeled water and indirect calorimetry, are complex, expensive, uncomfortable, and/or difficult to apply on real time. To overcome these drawbacks, the goal of this paper is to propose a model for computing EE in real time (minute-by-minute) from heart rate and accelerometer signals.

APPROACH

The proposed model, which consists of an original branched model, uses heart rate signals for computing EE on moderate to vigorous physical activities and a linear combination of heart rate and counts per minute for computing EE on light to moderate physical activities. Model parameters were estimated from a given data set composed of 53 subjects performing 25 different physical activities (light-, moderate- and vigorous-intensity), and validated using leave-one-subject-out. A different database (semi-controlled in-city circuit), was used in order to validate the versatility of the proposed model. Comparisons are done versus linear and nonlinear models, which are also used for computing EE from accelerometer and/or HR signals.

MAIN RESULTS

The proposed piecewise model leads to more accurate EE estimations ([Formula: see text], [Formula: see text] and [Formula: see text] J kg^-1 min^-1 and [Formula: see text], [Formula: see text], and [Formula: see text] J kg^-1 min^-1 on each validation database).

SIGNIFICANCE

This original approach, which is more conformable and less expensive than the reference methods, allows accurate EE estimations, in real time (minute-by-minute), during a large variety of physical activities. Therefore, this model may be used on applications such as computing the time that a given subject spent on light-intensity physical activities and on moderate to vigorous physical activities (binary classification accuracy of 0.8155).

Fog Computing and Edge Computing Architectures for Processing Data From Diabetes Devices Connected to the Medical Internet of Things

The Internet of Things (IoT) is generating an immense volume of data. With cloud computing, medical sensor and actuator data can be stored and analyzed remotely by distributed servers. The results can then be delivered via the Internet. The number of devices in IoT includes such wireless diabetes devices as blood glucose monitors, continuous glucose monitors, insulin pens, insulin pumps, and closed-loop systems. The cloud model for data storage and analysis is increasingly unable to process the data avalanche, and processing is being pushed out to the edge of the network closer to where the data-generating devices are. Fog computing and edge computing are two architectures for data handling that can offload data from the cloud, process it nearby the patient, and transmit information machine-to-machine or machine-to-human in milliseconds or seconds. Sensor data can be processed near the sensing and actuating devices with fog computing (with local nodes) and with edge computing (within the sensing devices). Compared to cloud computing, fog computing and edge computing offer five advantages: (1) greater data transmission speed, (2) less dependence on limited bandwidths, (3) greater privacy and security, (4) greater control over data generated in foreign countries where laws may limit use or permit unwanted governmental access, and (5) lower costs because more sensor-derived data are used locally and less data are transmitted remotely. Connected diabetes devices almost all use fog computing or edge computing because diabetes patients require a very rapid response to sensor input and cannot tolerate delays for cloud computing.

Future of Automated Insulin Delivery Systems

Advances in continuous glucose monitoring (CGM) have brought on a paradigm shift in the management of type 1 diabetes. These advances have enabled the automation of insulin delivery, where an algorithm determines the insulin delivery rate in response to the CGM values. There are multiple automated insulin delivery (AID) systems in development. A system that automates basal insulin delivery has already received Food and Drug Administration approval, and more systems are likely to follow. As the field of AID matures, future systems may incorporate additional hormones and/or multiple inputs, such as activity level. All AID systems are impacted by CGM accuracy and future CGM devices must be shown to be sufficiently accurate to be safely incorporated into AID. In this article, we summarize recent achievements in AID development, with a special emphasis on CGM sensor performance, and discuss the future of AID systems from the point of view of their input-output characteristics, form factor, and adaptability.

Moving Toward a Unified Platform for Insulin Delivery and Sensing of Inputs Relevant to an Artificial Pancreas

Advances in insulin pump and continuous glucose monitoring technology have primarily focused on optimizing glycemic control for people with type 1 diabetes. There remains a need to identify ways to minimize the physical burden of this technology. A unified platform with closely positioned or colocalized interstitial fluid glucose sensing and hormone delivery components is a potential solution. Present challenges to combining these components are interference of glucose sensing from proximate insulin delivery and the large discrepancy between the life span of current insulin infusion sets and glucose sensors. Addressing these concerns is of importance given that the future physical burden of this technology is likely to be even greater with the ongoing development of the artificial pancreas, potentially incorporating multiple hormone delivery, glucose sensing redundancy, and sensing of other clinically relevant nonglucose biochemical inputs.

Mitigating Reductions in Glucose During Exercise on Closed-Loop Insulin Delivery: The Ex-Snacks Study

OBJECTIVE

To assess whether snacking could be used with closed-loop (CL) insulin delivery to avoid exercise-induced reductions in plasma glucose (PG), as well as elevations in PG at the end of exercise.

RESEARCH DESIGN AND METHODS

Twelve type 1 diabetes (T1D) subjects (age 13-36 years, duration 10.7 ± 8.4 years, A1c 7.4% ± 0.8% [57 ± 8.7 mmol/mol]) underwent two 105-min exercise studies while under CL control: CL alone and CL+snack. Exercise, commenced at 3 PM, consisted of four 15-min periods of brisk treadmill walking to 65%-70% HR_max (separated by three 5-min rest periods), followed by a 30-min recovery period. Fifteen to 30 g carbohydrate (Gatorade) was provided on snacking visits just before and midway through the exercise period. PG and insulin were measured every 15-20 min during the exercise studies.

RESULTS

Baseline PG levels were similar for CL alone (164 ± 16 mg/dL) versus CL+snack (172 ± 11 mg/dL). During exercise, PG levels fell by 53 ± 10 mg/dL without snacking versus a modest 10 ± 13 mg/dL increase in PG with snacking (P = 0.0005); similar differences in the change in PG levels were observed at the end of recovery period. Hypoglycemia requiring rescue treatment (PG ≤60 mg/dL) during exercise occurred in three nonsnacking visits versus none with snacking. During the 75-min exercise period, insulin delivered was 1.8 ± 0.4 U for the CL+snack admission compared to 0.7 ± 0.1 U during CL alone (P = 0.002).

CONCLUSION

These results support the use of a simple snacking strategy to avoid exercise-induced lowering of PG while on CL insulin delivery. Persistent insulin infusion during exercise with snacking also appears to be effective in limiting increases in PG at the end of exercise.

Efficacy of single-hormone and dual-hormone artificial pancreas during continuous and interval exercise in adult patients with type 1 diabetes: randomised controlled crossover trial

AIMS/HYPOTHESIS

The aim of this study was to assess whether the dual-hormone (insulin and glucagon) artificial pancreas reduces hypoglycaemia compared with the single-hormone (insulin alone) artificial pancreas during two types of exercise.

METHODS

An open-label randomised crossover study comparing both systems in 17 adults with type 1 diabetes (age, 37.2 ± 13.6 years; HbA_1c, 8.0 ± 1.0% [63.9 ± 10.2 mmol/mol]) during two exercise types on an ergocycle and matched for energy expenditure: continuous (60% [Formula: see text] for 60 min) and interval (2 min alternating periods at 85% and 50% [Formula: see text] for 40 min, with two 10 min periods at 45% [Formula: see text] at the start and end of the session). Blocked randomisation (size of four) with a 1:1:1:1 allocation ratio was computer generated. The artificial pancreas was applied from 15:30 hours until 19:30 hours; exercise was started at 18:00 hours and announced 20 min earlier to the systems. The study was conducted at the Institut de recherches cliniques de Montréal.

RESULTS

During single-hormone control compared with dual-hormone control, exercise-induced hypoglycaemia (plasma glucose <3.3 mmol/l with symptoms or <3.0 mmol/l regardless of symptoms) was observed in four (23.5%) vs two (11.8%) interventions (p = 0.5) for continuous exercise and in six (40%) vs one (6.25%) intervention (p = 0.07) for interval exercise. For the pooled analysis (single vs dual hormone), the median (interquartile range) percentage time spent at glucose levels below 4.0 mmol/l was 11% (0.0-46.7%) vs 0% (0-0%; p = 0.0001) and at glucose levels between 4.0 and 10.0 mmol/l was 71.4% (53.2-100%) vs 100% (100-100%; p = 0.003). Higher doses of glucagon were needed during continuous (0.126 ± 0.057 mg) than during interval exercise (0.093 ± 0.068 mg) (p = 0.03), with no reported side-effects in all interventions.

CONCLUSIONS/INTERPRETATION

The dual-hormone artificial pancreas outperformed the single-hormone artificial pancreas in regulating glucose levels during announced exercise in adults with type 1 diabetes.

TRIAL REGISTRATION

ClinicalTrials.gov NCT01930110 FUNDING: : Société Francophone du Diabète and Diabète Québec.

Randomized trial of a dual-hormone artificial pancreas with dosing adjustment during exercise compared with no adjustment and sensor-augmented pump therapy

AIMS

To test whether adjusting insulin and glucagon in response to exercise within a dual-hormone artificial pancreas (AP) reduces exercise-related hypoglycaemia.

MATERIALS AND METHODS

In random order, 21 adults with type 1 diabetes (T1D) underwent three 22-hour experimental sessions: AP with exercise dosing adjustment (APX); AP with no exercise dosing adjustment (APN); and sensor-augmented pump (SAP) therapy. After an overnight stay and 2 hours after breakfast, participants exercised for 45 minutes at 60% of their maximum heart rate, with no snack given before exercise. During APX, insulin was decreased and glucagon was increased at exercise onset, while during SAP therapy, subjects could adjust dosing before exercise. The two primary outcomes were percentage of time spent in hypoglycaemia (<3.9 mmol/L) and percentage of time spent in euglycaemia (3.9-10 mmol/L) from the start of exercise to the end of the study.

RESULTS

The mean (95% confidence interval) times spent in hypoglycaemia (<3.9 mmol/L) after the start of exercise were 0.3% (-0.1, 0.7) for APX, 3.1% (0.8, 5.3) for APN, and 0.8% (0.1, 1.4) for SAP therapy. There was an absolute difference of 2.8% less time spent in hypoglycaemia for APX versus APN (p = .001) and 0.5% less time spent in hypoglycaemia for APX versus SAP therapy (p = .16). Mean time spent in euglycaemia was similar across the different sessions.

CONCLUSIONS

Adjusting insulin and glucagon delivery at exercise onset within a dual-hormone AP significantly reduces hypoglycaemia compared with no adjustment and performs similarly to SAP therapy when insulin is adjusted before exercise.

Nonadjunctive Use of Continuous Glucose Monitoring for Diabetes Treatment Decisions

While self-monitoring of blood glucose (SMBG) is the current standard used by people with diabetes to manage glucose levels, recent improvements in accuracy of continuous glucose monitoring (CGM) technology are making it very likely that diabetes-related treatment decisions will soon be made based on CGM values alone. Nonadjunctive use of CGM will lead to a paradigm shift in how patients manage their glucose levels and will require substantial changes in how care providers educate their patients, monitor their progress, and provide feedback to help them manage their diabetes. The approval to use CGM nonadjunctively is also a critical step in the pathway toward FDA approval of an artificial pancreas system, which is further expected to transform diabetes care for people with type 1 diabetes. In this article, we discuss how nonadjunctive CGM is expected to soon replace routine SMBG and how this new usage scenario is expected to transform health outcomes and patient care.

Reducing Glucose Variability Due to Meals and Postprandial Exercise in T1DM Using Switched LPV Control: In Silico Studies

BACKGROUND

Time-varying dynamics is one of the main issues for achieving safe blood glucose control in type 1 diabetes mellitus (T1DM) patients. In addition, the typical disturbances considered for controller design are meals, which increase the glucose level, and physical activity (PA), which increases the subject's sensitivity to insulin. In previous works the authors have applied a linear parameter-varying (LPV) control technique to manage unannounced meals.

METHODS

A switched LPV controller that switches between 3 LPV controllers, each with a different level of aggressiveness, is designed to further cope with both unannounced meals and postprandial PA. Thus, the proposed control strategy has a "standard" mode, an "aggressive" mode, and a "conservative" mode. The "standard" mode is designed to be applied most of the time, while the "aggressive" mode is designed to deal only with hyperglycemia situations. On the other hand, the "conservative" mode is focused on postprandial PA control.

RESULTS

An ad hoc simulator has been developed to test the proposed controller. This simulator is based on the distribution version of the UVA/Padova model and includes the effect of PA based on Schiavon.(1) The test results obtained when using this simulator indicate that the proposed control law substantially reduces the risk of hypoglycemia with the conservative strategy, while the risk of hyperglycemia is scarcely affected.

CONCLUSIONS

It is demonstrated that the announcement, or anticipation, of exercise is indispensable for letting a mono-hormonal artificial pancreas deal with the consequences of postprandial PA. In view of this the proposed controller allows switching into a conservative mode when notified of PA by the user.

Sensor Monitoring of Physical Activity to Improve Glucose Management in Diabetic Patients: A Review

Diabetic individuals need to tightly control their blood glucose concentration. Several methods have been developed for this purpose, such as the finger-prick or continuous glucose monitoring systems (CGMs). However, these methods present the disadvantage of being invasive. Moreover, CGMs have limited accuracy, notably to detect hypoglycemia. It is also known that physical exercise, and even daily activity, disrupt glucose dynamics and can generate problems with blood glucose regulation during and after exercise. In order to deal with these challenges, devices for monitoring patients’ physical activity are currently under development. This review focuses on non-invasive sensors using physiological parameters related to physical exercise that were used to improve glucose monitoring in type 1 diabetes (T1DM) patients. These devices are promising for diabetes management. Indeed they permit to estimate glucose concentration either based solely on physical activity parameters or in conjunction with CGM or non-invasive CGM (NI-CGM) systems. In these last cases, the vital signals are used to modulate glucose estimations provided by the CGM and NI-CGM devices. Finally, this review indicates possible limitations of these new biosensors and outlines directions for future technologic developments.