Comparative Analysis of Near-Infrared Light Absorption in Glucose and Blood Component Solutions With and Without 1 MHz Ultrasound Enhancement

Abstract

Diabetes is a global health issue that affects approximately 537 million people. The prevalence of diabetes is projected to reach 783 million by 2045. Invasive glucometers and patch sensors provide effective management through continuous monitoring, but these devices are inconvenient and lead to poor compliance. Alternative methods such as non-invasive near-infrared (NIR) monitoring offer continuous monitoring, but these devices are also often vulnerable to ambient light interference and absorption by substances such as ascorbic acid. Recent studies have shown that ultrasound can enhance optical interactions and increase measurement sensitivity in complex fluids. In this study, a proof-of-concept system combining 1 MHz ultrasound with a 940 nm NIR LED-photodiode circuit was developed. Ultrasoundassisted glucose detection was evaluated under controlled fluid-solution conditions at glucose concentrations ranging from 0 to 300 mg/dL. To assess selectivity, samples were tested with and without 30 mg/dL ascorbic acid under ultrasound and non-ultrasound conditions. The results showed that ultrasound increased the sensitivity from 113 µV/(mg·dL⁻¹) to 124 µV/(mg·dL⁻¹) in glucose-only samples and from 71 µV/(mg·dL⁻¹) to 74 µV/(mg·dL⁻¹) in glucose-ascorbic acid samples. The result corresponds to an increase insensitivity improvement from 7.37% to 19.60%. Ultrasound also improved linearity (R²) and reduced Root Mean Square Error (RMSE) across all test conditions, indicating improved consistency. In terms of selectivity, the measurement remained high (>98%), with voltage differences below 2% between glucose-only and glucose-ascorbic acid samples. This confirms the minimal influence of external factors under the tested conditions. Overall, these findings indicate that combining ultrasound and NIR for glucose detection can improve the sensitivity and reliability of the signals in water-based glucose solutions. Further validation, such as using tissue-mimicking phantoms and in vivo studies, is required before this approach can be considered for non-invasive glucose monitoring applications.