Linear System Identification Models for Non-Invasive Glucose Level Estimation

Abstract

Diabetes is a medical condition that can lead to severe health complications such as stroke, heart disease, blindness, and obesity. An estimated 347 million people worldwide were affected by diabetes, with approximately 3.4 million deaths attributed to high blood sugar levels. Researchers have explored various non-invasive techniques for measuring blood glucose levels, including ultrasonic sensors, multisensory systems, transmittance absorbance, bio-impedance, voltage intensity, and thermography.

This paper examines the application of near-infrared (NIR) spectroscopy for glucose level measurement and the implementation of a linear system identification model to predict output data from NIR measurements. While NIR has been utilized in previous studies, there is ongoing debate regarding the optimal wavelength range, as different researchers have used varying wavelengths. To assess the feasibility of a linear approach, this study applies the Autoregressive Moving Average Exogenous (ARMAX) model to predict NIR measurement outcomes