Non‐invasive glucose sensing methodologies for the measurement of human blood glucose levels

Abstract

This thesis explores the innovative field of non-invasive glucose monitoring (NGM), a promising avenue for improving diabetes management by providing painless and continuous monitoring of blood glucose levels. Diabetes, a chronic condition affecting millions globally, requires effective glucose monitoring to manage complications. Traditional invasive methods, such as finger-prick tests and continuous glucose monitors, although effective, often cause discomfort and compliance issues. A key contribution of this work is the development of a Time-of-Flight (ToF)- based sensor designed for glucose detection in physiological conditions. This novel optical sensor utilizing the Time-of-Flight principle was developed and successfully demonstrated its capability to achieve adequate depth penetration and sensitivity for detecting glucose-induced optical changes in human skin. Computational simulations,including Finite Element Method (FEM) modeling and Monte Carlo simulations, were employed to optimize the sensor design and predict light-tissue interactions. Despite its promising performance, the sensor’s accuracy was influenced by factors such as hydration levels, tissue scattering, and external environmental conditions. Efforts to measure glucose using bioimpedance spectroscopy through RF scattering parameters (S11 and S21) revealed repeatable but inconclusive correlations with glucose levels. The findings underscore the potential of optical methods while highlighting the need for multi-modal systems to address the inherent limitations of individual techniques. Future research directions include integrating optical and bioimpedance sensing modalities into a unified framework, leveraging machine learning for data fusion, and reducing environmental and physiological noise. These advancements aim to achieve robust, accurate, and clinically viable non-invasive glucose monitoring systems.