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.

Model based study on monitoring ketone bodies to improve safety in intensive insulin therapy.

AIM: The present study explores new signals (capillary 3betahydroxybutyrate - 3betaOHB) for improving the safety of a closed loop insulin infusion system (external wearable artificial pancreas). METHODS: Data collected during a clinical trial on diabetic subjects including a period of insulin deprivation were interpreted through mathematical models to simulate the effect of monitoring ketone bodies (capillary 3betaOHB, KB) compared to blood glucose in subjects on Continuous Subcutaneous Insulin Infusion (CSII) treatment. RESULTS: The estimation of model coefficients satisfactorily fits experimental data. The evaluation of dynamic changes of capillary 3betaOHB levels showed a more rapid response than blood glucose. CONCLUSIONS: The effect of the combination of monitoring of glucose and ketone bodies can consistently improve the safety and efficacy of the use of a closed loop system for glycemic control in dangerous situations like ketoacidosis. If a subcutaneous glucose-ketone bodies sensor were to become available in the near future it would be a key component of an external artificial pancreas.

A mathematical model describing the glycemic response of diabetic patients to meal and i.v. infusion of insulin.

Nine type 1 diabetic patients were studied for 24 hours. During this period they were given three calibrated meals. The glycemia was feedback-controlled by means of an artificial pancreas. The blood concentration of glucose and the infusion speed of the insulin were measured every minute. The experimental data referring to each of the three meals were used to estimate the parameters of a mathematical model suitable for describing the glycemic response of diabetic patients at meals and at the i.v. infusion of exogenous insulin. From the estimate a marked dispersion of the parameters was found, both interindividual and intraindividual. Nevertheless the models thus obtained seem to be usable for the synthesis of a feedback controller, especially in view of creating a portable artificial pancreas that now seems possible owing to the realization (so far experimental) of sufficiently reliable glucose concentration sensors.

A simulation study on a self-tuning portable controller of blood glucose.

A self-tuning, nonlinear controller was developed to drive portable or implantable micro-pumps for blood glucose control in diabetic subjects. The parameter estimation is based on a recursive least-square algorithm applied to a discrete time simplified mathematical model of the glucose system; the controller, which is integrated with the estimator, uses an extended minimum-variance method. The software was designed to drive a micro-controller wearable unit, operating with different kinds of glucose sensors, in order to perform chemical and biological experimentations. Some strategies were introduced to avoid hypoglycaemia, even tolerating a reduction in control speed and accuracy. The whole system was tested in a simulation study, performed on a mathematical model implemented on a personal computer. The tests were performed while simulating different controller structures and settings, and patient responses. They showed a satisfactory control behaviour, mostly as far as stability and robustness are concerned, in all simulated conditions.

Interferon-gamma (r-IFN-gamma) induced activation of alveolar macrophages (AM) from anergic patients with chronic obstructive pulmonary disease (COPD).

Forty-six anergic patients (37 males and 9 females, age range 55-79 yr) were selected from ninety-one patients suffering from COPD due to frequent exacerbations and impaired delayed cutaneous reactivity (43.9%). The phenotype of circulating lymphocytes, their proliferative response to a panel of polyclonal T-cell activators and the candidacidal activity (CA) of circulating PMNs (polymorphonuclear cells) were measured. In 13 patients presenting a defective CA of circulating PMNs, the in vitro response of alveolar macrophage CA to r-IFN-gamma was also determined. We found: 1) a significant reduction in the CL response to PHA in COPD patients vs controls; 2) a low PMN-CA in 23 (57%) COPD patients; 3) a non-significant difference in phenotype analysis in patients and controls; 4) lower CA of AMs in COPD patients than in controls; 5) restoration in vitro of CA by r-IFN-gamma in the group of anergic COPD patients presenting depressed CA. We conclude that a defective cell-mediated immunity could be the basis of the enhanced susceptibility to infectious exacerbations in many COPD patients and that, in vitro, it could be reversed by r-IFN-gamma treatment.

Time delay compensation for closed-loop insulin delivery systems: a simulation study.

Closed loop insulin therapy certainly represents the best possible approach to insulin replacement. However, present limitations preclude wider application of the so-called artificial pancreas. Therefore, a thorough understanding of these limitations is needed to design better systems for future long-term use. The present simulation study was design: to obtain better information on the impact of the measurement delay of currently available closed-loop devices both during closed-loop insulin delivery and blood glucose clamp studies, and to design and test a time delay compensator based on the method originally described by O.J. Smith. Simulations were performed on a Compaq Deskpro 486/25 personal computer under MS-DOS operating system using Simnon rel. 3.00 software. There was a direct relationship between measurement delay and amount of insulin delivered, i.e., the longer the delay the higher the insulin dose needed to control a rise in blood glucose; the closed-loop response in presence of a time delay was qualitatively impaired both during insulin delivery and blood glucose clamp studies; time delay compensation was effective in reducing the insulin dose and improving controller stability during the early phase of clamp studies. However, the robustness of a Smith's predictor-based controller should be carefully evaluated before implementation in closed-loop systems can be considered.

Wearable system for acquisition, processing and storage of the signal from amperometric glucose sensors.

A wearable device for the acquisition, processing and storage of the signal from needle-type glucose sensors has been designed and developed as part of a project aimed at developing a portable artificial pancreas. The device is essential to assess the operational characteristics of miniaturized sensors in vivo. It can be connected to sensors operating at a constant potential of 0.65 Volts, and generating currents in the order of 10(-9) Amp. It is screened and equipped with filters that permit data recording and processing even in the presence of electrical noise. It can operate with sensors with different characteristics (1-200 nA full scale). The device has been designed to be worn by patients, so its weight and size have been kept to a minimum (250 g; 8.5 x 14.5 x 3.5 cm). It is powered by rechargeable Ni/Cd batteries allowing continuous operation for 72 h. The electronics consists of an analog card with operational amplifiers, and a digital one with a microprocessor (Intel 80C196, MCS-96 class, with internal 16-bit CPU supporting programs written in either C or Assembler language), a 32 Kb EPROM, and an 8 Kb RAM where the data are stored. The microprocessor can run either at 5 or 10 Mhz and features on-chip peripherals: an analog/digital (A/D) converter, a serial port (used to transfer data to a Personal Computer at the end of the 72 h), input-output (I/O) units at high-speed, and two timers. The device is programmed and prepared to operate by means of a second hand-held unit equipped with an LCD display and a 16-key numeric pad.(ABSTRACT TRUNCATED AT 250 WORDS)

[A new proposal for nutritional education in a group of obese adolescents]. / Una nuova proposta di educazione alimentare in un gruppo di adolescenti obesi.

A new "active" system of monitoring and dietary education was experimented in a group of 20 obese adolescents waiting to start diet therapy. Each subject was requested to record all food and drink consumed during two separate periods, each lasting 7 days, at a distance of 15 days from each other, at home using an optic reader and book of bar codes. On the basis of answers to a questionnaire which was completed at the beginning and end of the study, aimed at assessing the level of knowledge of basic food hygiene, the majority of participants considered the experience useful and amusing and were willing to repeat it; in addition, a greater knowledge of nutritional principles and of the rations consumed was shown at the end of the study. The results of the study were analysed using the Food Meter-Miles computerised system and showed daily calorie intakes of 1514 +/- 524.0, M +/- DS (protein 18.3 +/- 5.1%, lipids 33.7 +/- 6.2%, glucose 47.9 +/- 8.6%, total fibre 14.4 +/- 4.6 g) in female subjects, and 2096.1 +/- 80.8 (protein 16.1 +/- 3.6%, lipids 38.2 +/- 2.9%, glucose 45.7 +/- 3.9%, total fibre 18.1 +/- 2.0 g) in male subjects. The number of foods chosen was very low considering the range of foods available (277) and the length of time studied: 35.7 +/- 11.8 and 35.0 +/- 6.0 respectively for female and male subjects. With regard to the number and type of meals eaten, a high number of snacks was observed which supplemented or replaced main meals.

Metabolic effects of intensified insulin therapy.

Intensified insulin therapy is usually carried out either with multiple subcutaneous insulin injections (ICT: intensified conventional therapy) or with continuous subcutaneous insulin infusion (CSII) by minipumps. For two years we have been studying two matched groups of type I diabetic patients, treated with 3 daily insulin injections (ICT) and with CSII ( Microjet , Miles), respectively. Blood glucose control, as assessed by integrated mean blood glucose (MBG), was similar in both groups, but a better metabolic stability ('M' index of Schlichtkrull ) was evident in the CSII group. From the 24-hour profiles of plasma 'free' IRI and metabolites (glucose, beta-OH-butyrate, lactate, pyruvate, alanine), both the hepatic and peripheral underinsulinization and related metabolic alterations were still evident in both groups of patients. The number of hypoglycaemic episodes, recorded by home blood glucose monitoring, was similar in both groups of patients, while the perception of symptomatic hypoglycaemia seemed to be reduced in the patients treated with CSII.