NBGM Model And Algorithm Research Based On PPG

With diabetes in the world growing day by day, real- time monitoring of blood glucose concentrations become necessary. In this paper, a NBGM Model and an intelligence algorithm of non- invasive blood glucose measuring is proposed. The main content of the paper is represented as followings:Firstly, NBGM System is proposed and the system model is set up. The entire system is divided into two parts, hardware system and software system.Secondly, to set up a hardware system of the NBGM system, including four major parts of the sensor, AFE4403, MSP430 Chip and a computer.Thirdly, proposing the intelligence algorithm of non- invasive blood glucose measuring, which is the software part of the N BGM. The algorithm includes Information acquiring, Features extracting, Machine Learning three mainly modules. Among above, Machine Learning module is the core part, according to the features extracting from the pulse wave, and then assembling them into a feature matrix. Putting the matrix and Glycemic Index into the machine learning module to train, thus gaining the predicting function.Fourthly, setting up experiments to compare and test the system module, hardware environment and software algorithm by computing correlation coefficients, standard deviations, Clarke error grid analysis, reliability verification system.At the end of the paper, we summarize the work of the paper and prospect the future of research.