Estimation Of Scattering Variation Insensitive Position And Its Application For Noninvasive Human Blood Glucose Measurement

Noninvasive blood glucose measurement(NBGM)based on near infrared spectroscopy is a promising method to continuously monitor blood glucose,so as it presents great significance for diabetic patients.In previous studies,we have found out a special source-detector separation(SDS),which can be used in NBGM to suppress the influence of scattering variation of human tissues,since there the diffuse light intensity only varies with tissue absorption change but does not vary with scattering change.This special SDS is called scattering variation independent SDS(SVI-SDS).However,the SVI-SDS is unknown for different tissues or individuals,and varying with wavelength.Therefore,it could not be easily applied for human tests since it is not be easily determined for tissues because we can not change the tissue’s scattering property by controllable approaches.This is a problem to be solved for its real application.In this thesis,we proposed a practical method to determine SVI-SDS for skin tissues by monitoring the differential diffuse reflection spectrum of tissue.This study would promote the real use of SVI-SDS for human blood glucose measurement.The main works in this thesis are as follows:1)The NBGM based on SVI-SDS is discussed.We simplified the used SDSs in NBGM to meet the requirement of different tissues and wavelengths.We use the measured diffuse light attenuance at two fixed SDSs to estimate the data for other SDSs including SVI-SDSs,because the attenuance changes with SDS linearly.2)A machine learning method for rapidly determining SVI-SDS is proposed.We use back propagation(BP)neural network to train the correlation between effective attenuation coefficient spectra of skin tissues and their SVI-SDSs.A prediction model for determining SVI-SDS is established.3)The SVI-SDS is applied on the real human NBGM.The determined SVI-SDSs of the tissue were used to monitor the skin’s absorption changes except the glucoserelated change.The skin changes were caused by some disturbances in the human test.Monte Carlo simulation and human experiments were performed.In conclusion,the work in this thesis may promote the real use of SVI-SDS into NBGM.Our approaches to determine and use the SVI-SDS are flexible and convenient in its practices.We applied SVI-SDS to monitor and correct the interference during test.we believe it is beneficial to improve the accuracy of NBGM.In addition,SVI-SDS can also be well applied to the other components’ measurement in scattering media.