A simulation-based study on the application of artificial neural networks to the NIR spectroscopic measurement of blood glucose

摘要

Diabetes Mellitus is a major health problem which affects about 200 million people worldwide.udDiabetics require their blood glucose levels to be kept within the normal range inudorder to prevent diabetes-related complications from occurring. Blood glucose measurementudis therefore of vital importance. The current glucose measurement techniques are, however,udpainful, inconvenient and episodic. This document provides an investigation into the useudof near-infrared spectroscopy for continuous, non-invasive measurement of blood glucose.udArtificial neural networks are used for the development of multivariate calibration modelsudwhich predict glucose concentrations based on the near-infrared spectral data. Simulationsudhave been performed which make use of simulated spectral data generated from the characteristicudspectra of many of the major components of human blood. The simulations showudthat artificial neural networks are capable of predicting the glucose concentrations of complexudaqueous solutions with clinically relevant accuracy. The effect of interference, such asudtemperature changes, pathlength variations, measurement noise and absorption due otherudanalytes, has been investigated and modelled. The artificial neural network calibrationudmodels are capable of providing acceptably accurate predictions in the presence of multipleudforms of interference. It was found that the performance of the measurement technique canudbe improved through careful selection of the optical pathlength and wavelength range for theudspectroscopic measurements, and by using preprocessing techniques to reduce the effect ofudinterference. Although the simulations suggest that near-infrared spectroscopy is a promisingudmethod of blood glucose measurement, which could greatly improve the quality of lifeudof diabetics, many further issues must be resolved before the long-term goal of developing audcontinuous non-invasive home glucose monitor can be achieved.