Determination of blood glucose concentration by back propagation neural network

Abstract

This paper presents a supervised backpropagation neural (BPN) network for the determination of blood glucose in diabetic patients. Non- invasive measurement of blood glucose concentration based on reflected laser beam from the index finger has been reported in this paper. This method depends on Helium-Neon (He-Ne) gas laser operating at 632.8 nm wave length. During measurement the index finger is placed in the laser beam transceiver unit, the reflected optical signal is converted into its corresponding electrical signal and the obtained signal is processed by the neural network which presents the results in the form of blood glucose concentration. Diabetes database used for empirical comparisons and the results are shown that BPN network performs better.