Search

The number of patients living with diabetes has increased significantly in recent years due to several factors. Many of these patients are choosing to use insulin pumps for their treatment, artificial systems that administer their insulin and consist of a glucometer and an automatic insulin supply working in an open loop. Currently, only a few closed-loop insulin delivery devices are commercially available. The most widespread systems among patients are what have been called the "Do-It-Yourself Hybrid Closed-Loop systems." These systems require the use of platforms with high computing power. In this paper, we will present a novel wearable system for insulin delivery that reduces the energy and computing consumption of the platform without affecting the computation requirements. Patients' information is obtained from a commercial continuous glucose sensor and a commercial insulin pump operating in a conventional manner. An ad-hoc embedded system will connect with the pump and the sensor to collect the glucose data and process it. That connection is accomplished through a radiofrequency channel that provides a suitable system for the patient. Thus, this system does not require to be connected to any other processor, which increases the overall stability. Using parameters configured by the patient, the control system will make automatic adjustments in the basal insulin infusion thereby bringing the patient's glycaemia to the target set by a doctor's prescription. The results obtained will be satisfactory as long as the configured parameters faithfully match the specific characteristics of the patient. Results from the simulation of 30 virtual patients (10 adolescents, 10 adults, and 10 children), using a python implementation of the FDA-approved (Food and Drug Administration) UVa (University of Virginia)/Padova Simulator and a python implementation of the proposed algorithm, are presented. [ABSTRACT FROM AUTHOR]