Non-Invasive Glucose Monitoring Using NIR Spectroscopy

Abstract

Abstract Diabetes Mellitus (DM), a metabolic condition, can lead to major health problems if it is not adequately treated. Traditional blood glucose monitors are intrusive and give people pain and irritation. As a result, the study’s goal was to find a non-invasive way to measure blood glucose. To determine blood glucose parameters, this system uses an NIR Spectroscopy-based system with a light source of wavelength 950nm. The light source illuminates the skin of the wrist, and a photodiode receiver captures the reflected beam. An Arduino UNO microcontroller receives the conditioned signal and digitalizes it. The spectrum is calculated by the Arduino board based on the subject’s blood glucose levels. Pre-processing and segmenting the raw waveforms are required to obtain the peaks of the NIR spectrum and to compute the regression model.