Personalised medicine for type 2 diabetic patient using neuro-fuzzy system
Wong, Chun Keet
Date of Issue2017
School of Computer Science and Engineering
Diabetes mellitus is a disease that affects many people, and by 2030, this figure is expected to hit 366 million, of which 90% or more are expected to be type 2 diabetes mellitus. The study presented in this report attempts to propose an automated insulin therapy, as opposed to the manual insulin bolus injections commonly used. A type 2 diabetes mellitus model is built using clinical data obtained to model the interactions of blood glucose concentration and insulin concentration. A new system is developed using a closed-loop regulator system which infuses insulin based on blood glucose concentration, and a neuro-fuzzy system, the eT2FIS, to predict insulin residual. The system monitors the patient’s blood glucose concentration, predict the insulin residual in the patient, and calculate the appropriate infusion rate for insulin. The system developed currently is still in the initial stages. The current system shows capability to regulate blood glucose levels and to keep them within a range. Further calibration of the system is required to regulate blood glucose concentration to the healthy range.
DRNTU::Engineering::Computer science and engineering
Final Year Project (FYP)
Nanyang Technological University