A closed-loop artificial pancreas based on model predictive control: Human-friendly identification and automatic meal disturbance rejection

Type 1 diabetes is characterized by a lack of insulin production by the pancreas, causing high blood glucose concentrations and requiring external insulin infusion to regulate blood glucose. Continuous glucose sensors can be coupled with continuous insulin infusion pumps to create a closed-loop artificial pancreas. A novel procedure of “human-friendly” identification testing using multisine inputs is developed to estimate suitable models for use in an artificial pancreas. A constrained model predictive control (MPC) strategy is developed to reduce risks of hypo- and hyperglycemia (low and high blood glucose concentration). Meal detection and meal size estimation algorithms are developed to improve meal glucose disturbance rejection when incoming meals are not announced. Closed-loop performance is evaluated through simulation studies of a type 1 diabetic individual, illustrating the ability of the MPC-based artificial pancreas control strategy to handle announced and unannounced meal disturbances.     

Neural network modeling and control of type 1 diabetes mellitus

This paper presents a developed and validated dynamic simulation model of type 1 diabetes, that simulates the progression of the disease and the two term controller that is responsible for the insulin released to stabilize the glucose level. The modeling and simulation of type 1 diabetes mellitus is based on an artificial neural network approach. The methodology builds upon an existing rich database on the progression of type 1 diabetes for a group of diabetic patients. The model was found to perform well at estimating the next glucose level over time without control. A neural controller that mimics the pancreas secretion of insulin into the body was also developed. This controller is of the two term type: one stage is responsible for short-term and the other for mid-term insulin delivery. It was found that the controller designed predicts an adequate amount of insulin that should be delivered into the body to obtain a normalization of the elevated glucose level. This helps to achieve the main objective of insulin therapy: to obtain an accurate estimate of the amount of insulin to be delivered in order to compensate for the increase in glucose concentration.     

Early post-prandial hyperglycemic response: from case reports of insulin-dependent diabetics evaluated with an artificial pancreas

In order to define the after-meal glycaemic response in diabetic subjects, we studied it in 20 insulin-dependent diabetic patients by automatic control using artificial beta-cell. We observed a different behaviour of blood glucose and of the consequent insulin requirement at the meals: in two subjects an early and excessive rise of blood glucose values at the meal was shown, not rapidly normalizable by artificial beta-cell. It should support, in our opinion, in these subjects, an increased neuro-entero-hormonal activity on endocrine pancreas production and a subsequent enhancement of hepatic glucose output.     

Model predictive control of glucose concentration in type I diabetic patients: An in silico trial

In this paper, the feedback control of glucose concentration in type I diabetic patients using subcutaneous insulin delivery and subcutaneous continuous glucose monitoring is considered. A recently developed in silico model of glucose metabolism is employed to generate virtual patients on which control algorithms can be validated against interindividual variability. An in silico trial consisting of 100 patients is used to assess the performances of a linear output feedback and a nonlinear state-feedback model predictive controller, designed on the basis of the in silico model. More than satisfactory results are obtained in the great majority of virtual patients. The experiments highlight the crucial role of the anticipative feedforward action driven by the meal announcement information. Preliminary results indicate that further improvements may be achieved by means of a nonlinear model predictive control scheme.     

Insulin substitution: new insulins, new modes of delivery

The demonstrated role of the tight control of hyperglycaemia for the prevention of long-term diabetic complications has reoriented the goals of insulin supply toward the search for restoration of the effects of physiological insulin secretion rather than the simple survival of insulin deficient patients and the reduction in the number of daily insulin injections to be performed. Normal blood glucose control requires the availability of a fast-acting insulin therapy at meal time in order to reduce hyperglycaemic excursions and a basal insulin therapy able to stabilize blood glucose between meals. Reduction of induced hypoglycaemic risk represents the secondary objective beside the main goal of avoiding hyperglycaemia. Fast-acting analogues, by a faster dissociation of their hexameric conformation after their injection or infusion in subcutaneous tissue, reduce post-meal hyperglycaemia, while their shortened duration of action versus regular insulin minimizes late post-absorptive risk of hypoglycaemia. Long-acting analogues, by their precipitation in subcutaneous tissue or their slowly reversible binding to albumin, provide a benefit on blood glucose stability versus NPH or zinc insulins. Continuous insulin therapy using pumps offers both a better blood glucose stability than multiple daily injections and a broader flexibility in life mode. Using the peritoneal route by implantable pumps is a mean to improve blood glucose stability in poorly controlled patients in spite of optimized subcutaneous insulin therapy. The development of glucose sensors provides reinforced information on blood glucose, versus self-monitoring by capillary blood measurements, that contributes to a better adaptation of insulin therapy. First trials of connections between blood glucose data and insulin delivery open a perspective toward glucose-modulated insulin therapy, at least in periods outside meals, leading to first models of semi-automated artificial endocrine pancreas. The alternative of a cellular insulin supply by pancreas or islet transplantation looked promising during recent years, but lack of transplants and adverse events related to immune suppression limit their use to very specific cases where benefit/risk ratio is positive.     

Can technological solutions for diabetes replace islet cell function?

The central objective of diabetes research and management is to restore the deficient secretion of insulin, thereby restoring a state of euglycemia and minimizing short- and long-term risks associated with poor glucose control. The development of the artificial pancreas seeks to imitate the action of the pancreatic beta cell by employing closed-loop control to respond to glycemic excursions by appropriately infusing appropriate amounts of insulin. This article examines progress towards implementing an artificial pancreas in the context of the pancreatic islet as the ideal model for controlling blood glucose. Physiologic insulin secretion will form our foundation for considering the technical design elements relevant to electromechanically imitating the beta cell. The most recent clinical trials using closed-loop control are reviewed and this modality is compared to other curative approaches including islet cell transplantation and preservation. Finally, the potential of the artificial pancreas as a method to adequately reestablish euglycemia is considered.     

A coordinated control strategy for insulin and glucagon delivery in type 1 diabetes

Type 1 diabetes is an autoimmune condition characterised by a pancreatic insulin secretion deficit, resulting in high blood glucose concentrations, which can lead to micro- and macrovascular complications. Type 1 diabetes also leads to impaired glucagon production by the pancreatic α-cells, which acts as a counter-regulatory hormone to insulin. A closed-loop system for automatic insulin and glucagon delivery, also referred to as an artificial pancreas, has the potential to reduce the self-management burden of type 1 diabetes and reduce the risk of hypo- and hyperglycemia. To date, bihormonal closed-loop systems for glucagon and insulin delivery have been based on two independent controllers. However, in physiology, the secretion of insulin and glucagon in the body is closely interconnected by paracrine and endocrine associations. In this work, we present a novel biologically-inspired glucose control strategy that accounts for such coordination. An in silico study using an FDA-accepted type 1 simulator was performed to evaluate the proposed coordinated control strategy compared to its non-coordinated counterpart, as well as an insulin-only version of the controller. The proposed coordinated strategy achieves a reduction of hyperglycemia without increasing hypoglycemia, when compared to its non-coordinated counterpart.     

Using mouse models to understand normal and abnormal urogenital tract development

The central objective of diabetes research and management is to restore the deficient secretion of insulin, thereby restoring a state of euglycemia and minimizing short- and long-term risks associated with poor glucose control. The development of the artificial pancreas seeks to imitate the action of the pancreatic beta cell by employing closed-loop control to respond to glycemic excursions by appropriately infusing appropriate amounts of insulin. This article examines progress towards implementing an artificial pancreas in the context of the pancreatic islet as the ideal model for controlling blood glucose. Physiologic insulin secretion will form our foundation for considering the technical design elements relevant to electromechanically imitating the beta cell. The most recent clinical trials using closed-loop control are reviewed and this modality is compared to other curative approaches including islet cell transplantation and preservation. Finally, the potential of the artificial pancreas as a method to adequately reestablish euglycemia is considered.     

The use of the artificial pancreas in uremic diabetic patients.

Maintenance hemodialysis and renal transplantation are increasingly used for treating diabetic patients with end-stage renal failure. The use of the artificial pancreas is able to prevent large blood glucose fluctuations in these patients with atherosclerosis, advanced retinopathy or neuropathy in which hyper- and hypoglycemia are potentially deleterious. For this purpose, we have developed and are utilizing an artificial pancreas easily utilizable without special training by the staff of a dialysis unit. This artificial pancreas uses a polarographic glucose electrode with a fast response time (45 to 90 seconds), a terminal display for operator communication, and a continuous digital and analogyl display for control of the running operation. There is also a printer to display in tabular and graphical form the values at any time during the operation. In this preliminary study, 7 patients have been studied: five under repetitive hemodialysis for four hours, 3 times a week; one treated by peritoneal dialysis for 12 hours, twice a week and one controlled during, and 48 hours after, renal transplantation. The macroscopic pancreas normalizes blood glucose under these circumstances, helps in a better understanding of blood glucose homeostasis in uremic patients under dialysis, leads to a more precise evaluation of insulin needs, may help to improve the nutritional status of the patients, and has an educational value for the patient and the medical staff.     

Blood glucose control algorithms for type 1 diabetic patients: A methodological review

A method for optimal continuous insulin therapy for diabetes patients has been sought since the early 1970s. Although technical and medical advances have been made, a fully automated artificial pancreas to replace the functions of the natural organ is still a research aim. This review compares recent control algorithms for type 1 diabetic patients which automatically connect continuous glucose monitoring and insulin injection, without patient intervention. Black-box model and gray-box model based control strategies are described and their performances are evaluated, with a focus on their feasibility of implementation in a real-life situation. In conclusion, a satisfactory control strategy has not yet been proposed, mainly because most control algorithms rely on continuous blood glucose measurement which is not yet available. Modeling the effect of glucose ingestion as an external disturbance on the time evolution of blood glucose concentration, is now the norm for the control community. In contrast, the effects of physical activity on the metabolic system is not yet fully understood and remain an open issue. Moreover, clinical studies on evaluation of control performance are scarce. Therefore, research on blood glucose control needs to concentrate on advanced patient modeling, control optimization and control performance evaluation under realistic patient-oriented conditions.     

Real-time estimation of plasma insulin concentration from continuous glucose monitor measurements

Continuous glucose monitors can measure interstitial glucose concentration in real time for closed-loop glucose control systems, known as artificial pancreas. These control systems use an insulin feedback to maintain plasma glucose concentration within a narrow and safe range, and thus to avoid health complications. As it is not possible to measure plasma insulin concentration in real time, insulin models have been used in literature to estimate them. Nevertheless, the significant inter- and intra-patient variability of insulin absorption jeopardizes the accuracy of these estimations. In order to reduce these limitations, our objective is to perform a real-time estimation of plasma insulin concentration from continuous glucose monitoring (CGM). Hovorka’s glucose–insulin model has been incorporated in an extended Kalman filter in which different selected time-variant model parameters have been considered as extended states. The observability of the original Hovorka’s model and of several extended models has been evaluated by their Lie derivatives. We have evaluated this methodology with an in-silico study with 100 patients with Type 1 diabetes during 25 h. Furthermore, it has been also validated using clinical data from 12 insulin pump patients with Type 1 diabetes who underwent four mixed meal studies. Real-time insulin estimations have been compared to plasma insulin measurements to assess performance showing the validity of the methodology here used in comparison with that formerly used for insulin models. Hence, real-time estimations for plasma insulin concentration based on subcutaneous glucose monitoring can be beneficial for increasing the efficiency of control algorithms for the artificial pancreas.     

Postprandial blood glucose control using a hybrid adaptive PD controller with insulin-on-board limitation

This paper addresses the design of blood glucose control during the postprandial period for Type 1 diabetes patients. An artificial pancreas for ambulatory purposes has to deal with the delays inherent to the subcutaneous route, the carbohydrate intakes, the metabolic changes, the glucose sensor errors and noise, and the insulin pump constraints. A time response typically obtained in closed-loop insulin delivery shows hyperglycemia in the early postprandial period caused by the lag in the insulin absorbtion, followed by hypoglycemia caused by control over-reaction. A hybrid control system is proposed in this paper to overcome these problems. An insulin bolus is administered prior to the meals like in open-loop control, whereas a PD controller is used for robust glucose regulation. The controller gain is progressively increased after the bolus from zero up to its nominal value as function of the insulin on board, so that the PD controller becomes fully operational just when the insulin on board falls below a prescribed value. An excessive accumulation of active insulin is avoided in this way, drastically reducing the risk of hypoglycemia. The controller gain is adapted by means of a variable structure algorithm, allowing a very simple software implementation. The robust performance of the control algorithm is intensively assessed in silico on a cohort of virtual patients under challenging realistic scenarios considering mixed meals, circadian variations, time-varying uncertainties, discrete measurement and actuation, sensor errors and other disturbances.     

An extension to the compartmental model of type 1 diabetic patients to reproduce exercise periods with glycogen depletion and replenishment

The purpose of this work is to present the main interactions promoted by exercise and synthesize them into mathematical equations. It is intended to extend the ability of the compartmental glucose-insulin model introduced by Sorensen [1985. A physiologic model of glucose metabolism in man and its use to design and assess improved insulin therapies for diabetes. Ph.D. Dissertation, Chemical Engineering Department, MIT, Cambridge] to reproduce variations in the blood glucose concentration induced by exercise in diabetic patients and to complement the previous work by Lenart and Parker [2002. Modeling exercise effects in type I diabetic patients. In: Proceedings of the 15th Triennial World Congress, Barcelona, Spain] and Lenart, DiMascio and Parker [2002. Modeling glycogen-exercise interactions in type I diabetic patients. In: Proceedings of the A.I.Ch.E. Annual Meeting, Indianapolis, IN]. The immediate consequences of exercise are incorporated in this research: redistribution of blood flows, increments in peripheral glucose and insulin uptakes, and increment in hepatic glucose production. The extended model was verified with experimental data for light and moderate intensity exercise. In addition, data extrapolation was introduced to simulate heavy intensity exercise. The hepatic glycogen reservoir limits the peripheral glucose uptake for prolonged exercise. Therefore, the depletion and replenishment of hepatic glycogen were modeled, looking to reproduce the blood glucose levels for a type 1 diabetic patient during a normal day, with meal intakes, insulin infusions and/or boluses, and a predefined exercise regime. From the extensive simulation evaluation, it is found that the new exercise model provides a good approximation to the available experimental data from literature.     

The effect of three days of blood glucose normalization by means of an "artificial endocrine pancreas" on the concentrations of growth hormone, glucagon, and cortisol in juvenile diabetics.

Glucagon, growth hormone, and cortisol secretion was studied in seven male insulin-dependent diabetics under conventional subcutaneous insulin therapy and after three days of blood glucose normalization attained by the artificial endocrine pancreas (Biostator-GCIIS). The diurnal hormonal profiles under the two types of therapy were compared. Six healthy male students served as control group. A three-day period of blood glucose normalization in insulin-dependent diabetic can restore glucagon secretion to normal. Growth hormone secretion is decreased but not completely normalised. Cortisol secretion is slightly decreased. It is concluded that prolonged normoglycemia achieved by means of an artificial endocrine pancreas may completely control endocrine abnormalities in insulin-dependent diabetics.     

Artificial pancreas: study methods and control of glycemia equilibrium in diabetic patients

The artificial endocrine pancreas or artificial beta cell, consists of a continuous blood glucose monitoring system, a computered set-up which responds to glycemia and the hormone delivery system. This system provides a technique which can completely normalize blood glucose concentration in diabetic patients, during both food assumption and after. This paper deals with the characteristics of the artificial beta cell, its possible applications and its limits.     

An iterative learning strategy for the auto-tuning of the feedforward and feedback controller in type-1 diabetes

This paper proposes a scheme for the control of the blood glucose in subjects with type-1 diabetes mellitus based on the subcutaneous (s.c.) glucose measurement and s.c. insulin administration. The tuning of the controller is based on an iterative learning strategy that exploits the repetitiveness of the daily feeding habit of a patient. The control consists of a mixed feedback and feedforward contribution whose parameters are tuned through an iterative learning process that is based on the day-by-day automated analysis of the glucose response to the infusion of exogenous insulin. The scheme does not require any a priori information on the patient insulin/glucose response, on the meal times and on the amount of ingested carbohydrates (CHOs). Thanks to the learning mechanism the scheme is able to improve its performance over time. A specific logic is also introduced for the detection and prevention of possible hypoglycaemia events. The effectiveness of the methodology has been validated using long-term simulation studies applied to a set of nine in silico patients considering realistic uncertainties on the meal times and on the quantities of ingested CHOs.     

Dynamic study of the blood viscosity of insulin-dependent diabetics by means of an artificial pancreas.

The blood viscosity of 15 insulin-dependent, poorly controlled, diabetic subjects was determined by using a microviscosimeter at low shear rates, with cylindrical cuvettes of the Couette type. It was found that the blood viscosity of these diabetics was more elevated than that of control patients (p less than 0.001). In ten diabetics, the return to a strict metabolic control over a period of more than 24 h by means of an artificial pancreas resulted in a significant systematic lowering of blood viscosity. These results suggest that the metabolic control of diabetes influences blood viscosity, and they underline the importance of an artificial pancreas for the dynamic study of the factors affecting blood viscosity during the course of diabetes.     

定期血糖监测对糖尿病患者血糖控制及生活方式的影响

目的:探讨定期血糖监测对糖尿病血糖控制及生活方式的影响。方法:随机抽取我中心2010-2011年度确诊的老年2型糖尿病患者110例,随机分为干预组和对照组,每组各55例;2组均接受正规降糖药物治疗及生活方式指导,干预组每周进行一次血糖监测,每3个月测一次糖化血红蛋白,对照组按患者意愿测定血糖指标,通过12个月的观察,研究两组患者在血糖控制及生活方式上的差异。结果:干预组患者空腹血糖(FPG)由定期监测血糖前的(7.26±1.36)mmol/L降至(6.68±1.10)mmol/L;餐后2小时血糖(2HPG)由定期监测血糖前的(12.34±2.29)mmol/L降至(11.09±1.98)mmol/L;糖化血红蛋白由监测前的(7.99±1.61)%降至(6.60±0.87)%;差异具有显著性(P<0.05);生活方式亦有明显改善,差异具有显著性(P<0.05);而对照组的改变不如干预组。结论:通过定期血糖监测可以有效地控制血糖、糖化血红蛋白,促使老年2型糖尿病患者改变不良生活方式。     

D-Trp8-D-Cys14-somatostatin--demonstration of its differential suppressive activity in juvenile diabetics.

In 8 insulin-dependent diabetics, the effect of D-Trp8-D-Cys14-somatostatin on blood glucose, growth hormone, and glucagon levels as well as on insulin requirements from an artificial endocrine pancreas was studied during a balanced meal. The somatostatin analogue was infused at a rate of 25 microgram/h preceeded by a bolus injection of 25 microgram 30 minutes before ingestion of the meal. At this dose the analogue had no effect on glucagon levels and insulin requirements from the artificial pancreas. On the other hand, there was a significant lowering effect on fasting blood glucose levels, possibly indicating a direct inhibition of hepatic glucose production. Furthermore, there might be a slight effect on growth hormone levels, as was demonstrated by a rebound increase after termination of analogue infusion.