Eduardo Ruiz-Velazquez

Weighting Restriction for Intravenous Insulin Delivery on T1DM Patient via

A weighting restriction with frequency components is proposed for the insulin delivery on Type 1 Diabetics Mellitus (T1DM) towards the control of the blood glucose level. The weighting restriction is stated from a model of healthy subjects which includes a rate for insulin delivery. The frequency components are incorporated via a transfer function from the plasma glucose to the free-plasma insulin such that a H infin-based controller is designed. In this way, the control synthesis involves the frequency components on which a healthy pancreas delivers insulin for the glucose homeostasis. In order to test controller performance, a dynamical model of an actuator is also included in the closed-loop system to add its effects in the closed-loop evaluation of the H infin -based controller. The actuator is a pump to deliver of an insulin infusion according with the rate computed by the controller. Note that the contribution is particularly focused on T1DM; however, the inclusion of weighting restriction can be used also onto critical care conditions.

H_{\infty}

Abstract In this paper, inverse optimal neural control for trajectory tracking is applied to glycemic control of type 1 diabetes mellitus (T1DM) patients. The proposed control law calculates the adequate insulin delivery rate in order to prevent hyperglycemia and hypoglycemia levels in T1DM patients. Two models are used: (1) a nonlinear compartmental model in order to obtain type 1 diabetes mellitus virtual patient behavior, and (2) a neural model obtained from an on-line neural identifier, which uses a recurrent neural network, trained with the extended Kalman filter (EKF); the last one allows the applicability of an inverse optimal neural controller. The proposed algorithm is tuned to track a desired trajectory; this trajectory reproduces the glucose absorption of a healthy person. The applicability of the proposed control scheme is illustrated via simulations.

Control

This paper deals with the blood glucose level modeling for Type 1 Diabetes Mellitus (T1DM) patients. The model is developed using a recurrent neural network trained with an extended Kalman filter based algorithm in order to develop an affine model, which captures the nonlinear behavior of the blood glucose metabolism. The goal is to derive a dynamical mathematical model for the T1DM as the response of a patient to meal and subcutaneous insulin infusion. Experimental data given by continuous glucose monitoring system is utilized for identification and for testing the applicability of the proposed scheme to T1DM subjects.

Inverse optimal neural control of blood glucose level for type 1 diabetes mellitus patients

The design of reduced-order controllers under specific performance and structure requirements is dealt in this contribution. Two controllers are designed and compared. The first one was designed using H∞ theory whereas the latter one is designed departing from a parametric-optimisation via a two-stage algorithm. The time spent by the designer using our second approach is largely reduced. An active suspension system is selected as a case study. The performance of both controllers is tested experimentally in the active suspension set-up. The experimental results show that the parametric-optimisation controller practically meets the desired performance specifications. Meanwhile, the H∞ controller cannot accomplish the imposed constraints just in the low-frequency range.

BLOOD GLUCOSE LEVEL NEURAL MODEL FOR TYPE 1 DIABETES MELLITUS PATIENTS

In this paper, discrete time inverse optimal trajectory tracking for a class of non-linear positive systems is proposed. The scheme is developed for MIMO (multi-input, multi-output) a!ne systems. This approach is adapted for glycemic control of type 1 diabetes mellitus (T1DM) patients. The control law calculates the insulin delivery rate in order to prevent hyperglycemia levels. A neural model is obtained from an on-line neural identifier, which uses a recurrent neural network, trained with the extended Kalman filter (EKF); this neural model has an a!ne form, which permits the applicability of inverse optimal control scheme. The proposed algorithm is tuned to follow a desired trajectory; this trajectory reproduces the glucose absorption of a healthy person. Simulation results illustrate the applicability of the control law in biological processes.

Design of Reduced-Order Controllers via H∞ and Parametric Optimisation: Comparison for an Active Suspension System

This paper presents on-line blood glucose level modeling for Type 1 Diabetes Mellitus (T1DM) patients. The model is developed using a recurrent neural network trained with an extended Kalman filter based algorithm in order to develop an affine model, which captures the nonlinear behavior of the blood glucose metabolism. The goal is to derive an on-line dynamical mathematical model of the T1DM for the response of a patient to meal and subcutaneous insulin infusion. Simulation results are utilized for identification and for testing the applicability of the proposed scheme.

Inverse optimal trajectory tracking for discrete time nonlinear positive systems

This paper deals with subcutaneous blood glucose level control. Inverse optimal trajectory tracking for discrete time non-linear positive systems is applied. The scheme is developed for MIMO (multi-input, multi-output) affine systems. The control law calculates the subcutaneous insulin delivery rate in order to prevent hyperglycemia and hypoglycemia events. A neural model is obtained from an on-line neural identifier, which uses a recurrent neural network, trained with the extended Kalman filter (EKF); this neural model has an affine form, which permits the applicability of inverse optimal control scheme. The proposed algorithm is tuned to follow a desired trajectory; this trajectory reproduces the glucose absorption of a healthy person. Then this model is used to synthesize an inverse optimal controller in order to regulate the subcutaneous blood glucose level for a Type 1 Diabetes Mellitus patient the applicability of the proposed scheme is illustrated via simulation using a recurrent neural network in ...

Neural model of blood glucose level for Type 1 Diabetes Mellitus Patients

Type 1 Diabetes mellitus (T1DM) is a chronic disease that occurs when the body cannot produce insulin. Since insulin was discovered in 1920, the way to keep T1DM patients blood glucose at normal levels has been insulin injections, via subcutaneous or intravenous paths. The efforts for an external infusion therapy have resulted in the so-called Artificial Pancreas. Such device attempts to integrate continuous insulin infusion, continuous glucose monitoring and an automatic control algorithm, which calculates the required insulin infusion. Considering all the problems related to T1DM, in this paper a neural model which captures the nonlinear behavior of the complex glucose-insulin dynamics is proposed; based on this model, a control algorithm is developed using the neural inverse optimal control via control lyapunov function (CLF) technique. Simulation results illustrate the applicability of the propounded scheme.

Neural Inverse Optimal Control via Passivity for Subcutaneous Blood Glucose Regulation in Type 1 Diabetes Mellitus Patients

This paper deals with blood glucose level control. Inverse optimal trajectory tracking via control Lyapunov function for discrete time non-linear systems is applied. The control law calculates the subcutaneous insulin delivery rate in order to prevent hyperglycemia and hypoglycemia levels. For this paper, a quadratic candidate CLF is used to synthesize the inverse optimal control law. The proposed algorithm is tuned to follow a desired trajectory; this trajectory reproduces the glucose absorption of a healthy person. Simulation results illustrate the applicability of the control law.

Subcutaneous neural inverse optimal control for an Artificial Pancreas

This paper presents the application of a recurrent multilayer perceptron neural network for modeling blood glucose dynamics in Type 1 Diabetes Mellitus (T1DM). Training is performed based on an extended Kalman filtering (EKF) learning algorithm. Then, the EKF performance is compared with the well-known Levenberg-Marquardt (LM) learning algorithm. The goal is to derive a dynamical mathematical model for T1DM considering the response of a patient to meal and subcutaneous insulin infusion. Thus, the main contribution of this work is to propose a modeling methodology for blood glucose dynamics based in Artificial Neural Networks (ANN). Experimental data, given by a continuous glucose monitoring system, are utilized for identification purposes and for applicability trials of the proposed scheme in T1DM therapy.