Non-Invasive Techniques With Vital Signs for Glucose Monitoring

Abstract

This study is focusing on integrating finger sleeves for machine-learning with Near-infrared (NIR) spectroscopy and additional sensors techniques. It applied light emitting diodes (LEDs) at <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$660 \text{nm}, 880 \text{nm}, 940 \text{nm}$</tex> wavelengths and photodetectors and a galvanic skin response (GSR) and a temperature sensor to read the signal from patients' fingers. These sensors are attached to the finger sleeves to make it easy to wear for this continuous glucose monitoring. After the data was collected from the NIR spectroscopy and multiple sensors it has used in the machine learning models to predict the blood sugar level. For the machine learnings that was selected in this study are Linear regression, and Random forest model. Which the <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{R}^{\mathbf{2}}$</tex> result was 0.07 and -0.27 respectively.