Blood Glucose Level Prediction Using Intelligent Based Modeling Techniques

In this paper, new methods of blood glucose (BG) level modeling and prediction using artificial intelligent (AI) based techniques have been suggested. The obtained data was windowed and normalized prior to the application of AI based modeling techniques for the estimation of appropriate model parameters. The obtained parameters can be used to predict the next pre-breakfast BG level. Performance analysis of the suggested techniques have been evaluated using pre-breakfast BG level data and results obtained show that the techniques especially the hybrid AIS technique can accurately predict next BG level, thus critical emergency can be avoided and at the same time injection of insulin dosages and appropriate meals for a diabetic patient can be prepared ahead of time.