MODELING AND SIMULATION OF FUZZY BASED AUTOMATIC INSULIN DELIVERY SYSTEM

Abstract

The aim is to design a complete online, automatic, intelligent and independent insulin delivery system. The proposed system calculates the amount of insulin by formulation based upon two linguistic factors i.e. weight and blood glucose levels. These factors have been used to develop a rule based fuzzy system in LabVIEW. The fuzziness brings a better regulation into the regime than a non fuzzy system as small changes in quantities of insulin can bring about effective command on the glucose levels. A user friendly interactive webpage has been designed that inculcates the fuzzy system for providing online doctor consultation. The insulin delivery has been simulated in LabVIEW where the quantity of insulin delivered to the patient can be controlled automatically or by the doctor. The system has been verified by taking a random sample of hundred insulin dependent individuals to test the effectiveness of the system. The developed fuzzy system was found to be more accurate than a non fuzzy system. This study is modeled for soluble human insulin type. The implementation of this system in real time with insulin pump will enable those in inaccessible areas and intensive care units to control the glucose levels automatically in an efficient manner.