Neural network based blood glucose level prediction in type I diabetes patients
Engineering Education # 07, July 2011
The research was a part of the project aimed at developing an automated optimal insulin dose forecasting system, which, when combined with an insulin pump and continuous glucose monitoring system, could serve as an artificial pancreas. Paper investigates the effectiveness of neural network algorithms for predicting the blood glucose values in type I diabetes patients. The aim is to choose the optimal neural network and its training algorithm for use in predicting optimal doses of insulin. This research was performed using a software system MatLab.