Performance of optical sensing for blood glucose measurement

Pao Yue-kong Library Electronic Theses Database

Performance of optical sensing for blood glucose measurement

 

Author: So, Chi Fuk
Title: Performance of optical sensing for blood glucose measurement
Degree: Ph.D.
Year: 2013
Subject: Blood -- Analysis.
Optical detectors.
Hong Kong Polytechnic University -- Dissertations
Department: School of Nursing
Pages: ix, 164 leaves : ill. (some col.) ; 30 cm.
Language: English
InnoPac Record: http://library.polyu.edu.hk/record=b2639075
URI: http://theses.lib.polyu.edu.hk/handle/200/7136
Abstract: Diabetes mellitus is an intractable condition in which blood glucose levels cannot be regulated normally by the body alone; it has many complications, including heart disease and stroke, kidney failure, blindness or vision problems, diabetic neuropathy and diabetic foot. Treatment methods include dietary regulation to control blood glucose levels, oral medication, and insulin injection, and all of these treatments should rely on blood glucose measurement. Diabetes mellitus was the tenth most common cause of deaths in Hong Kong in 2010. Currently, the most common means of checking is by using a finger prick glucose meter, but many people dislike using sharp objects and seeing blood because there is a risk of infection. Over the long term, this practice may also result in damage to finger tissue. Given these realities, the advantages of a non-invasive technology are easily understood. The race for the next generation of painless and reliable glucose monitoring for diabetes is on. As technology advances, so diagnostic techniques and equipment improve. Near infrared (NIR) spectroscopy has become a promising technology, among others, for blood glucose monitoring. While advances have been made, the reliability and the calibration of non-invasive instruments could still be enhanced, and the search has continued to the present without a clinically or commercially viable product emerging. The aim of this study is to evaluate a self-monitoring medical device adopting NIR spectroscopy, which is able to detect glucose concentrations non-invasively. Moreover, the precision in blood glucose measurement will also be validated. The objective of this study was to set up a non-invasive blood glucose measurement device that is stable and easy to use to detect the spectral response from human tissue. It was then used to examine different locations of the human body for non-invasive measurement, and the temperature difference of human tissues was inspected. In addition, various pre-processing methods were compared and a series of NIR wavelengths was identified. After that, robust mathematical models for classification and regression approaches were constructed that are able to classify and predict glucose concentrations in blood vessels non-invasively.
Partial least squares (PLS) is widely used in multivariate calibration methods. Partial least squares discriminant analysis (PLS-DA) is a variant of PLS when the dependent variable is binary. They are particularly useful in spectral analysis because the concurrent inclusion of large spectral data for the analyte can greatly improve the precision and applicability of multivariate analysis. Very often, only one single quantitative model is constructed to predict the relationship between the response and the independent variables. This approach can easily misidentify, under or over estimate the important features contained in the independent variables. The results obtained by a single prediction model are thus unstable or correlated to spurious spectral variance, particularly when the training set for PLS is relatively small. New algorithms developed by applying the Monte Carlo (MC) method to PLS and PLS-DA, namely MC-PLS and MC-PLS-DA respectively, are proposed to classify spectral data obtained from NIR blood glucose measurement. Noise in the data is removed by randomly selecting different subsets from the whole training dataset to generate a large number of models. The new algorithms are then used in determining the mean value over the models with high correlation and small prediction errors for MC-PLS, or the mean sensitivity and specificity of these models are then calculated to determine the model with the best classification rate for MC-PLS-DA. The results show that both the MC-PLS and MC-PLS-DA methods give more accurate prediction results when compared with other multivariate methods used for NIR spectroscopic data of blood glucose. Additionally, the stability of the MC-PLS and MC-PLS-DA models are enhanced compared with the conventional PLS and PLS-DA models. The MC-PLS and the MC-PLS-DA methods are proposed in this study to tackle the problems in which accuracy is limited by the use of one single prediction model. These methods integrate the Monte Carlo method into the conventional PLS and PLS-DA to improve performance. The proposed algorithms exhibit better performance and accuracy rates when compared to other multivariate methods, as evident from the prediction results on the NIR spectral data. The prediction of the relationship between the response and the independent variables is more accurate, thus enhancing the reliability of the regression model. These advantages make MC-PLS and MC-PLS-DA a promising approach for non-invasive estimation of blood glucose.

Files in this item

Files Size Format
b26390759.pdf 3.018Mb PDF
Copyright Undertaking
As a bona fide Library user, I declare that:
  1. I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
  2. I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
  3. I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.

     

Quick Search

Browse

More Information