Non-invasive glucose measurement using 950 nm reflective short wave NIR technique

Purpose: Diabetes is one of the most common diseases in the world. A finger prick is typically required to obtain blood samples for diabetes testing. It is uncomfortable and prone to infection to undergo these treatments. To address this issue, a non-invasive short wave near-infrared-based optical detection system with a 950 nm wavelength sensor operating in a reflective mode is proposed. This research seeks to improve the accuracy of blood glucose detection non-invasively with low cost. Methods: To estimate glucose, a 950 nm reflective sensor is used, and the measured signal is converted to voltages and passed through a precision analog-to-digital converter. The relationship between voltage and predicted glucose is evaluated using absorbance and transmittance measurements. The proposed method employs a third-order polynomial expression to improve accuracy through the use of an algorithm. In total, 184 real-time samples were collected using the reference finger prick glucose device, of which 25 were collected while fasting and the remaining during non-fasting periods. The glucose evaluation metrics are computed and surveillance error grid (SEG) analysis is carried out. Results: The proposed technique utilizing a 950 nm reflective sensor yields efficient results compared to existing literature. Real-time data analysis using the proposed sensor method shows an increase in the coefficient of determination (R 2) value and mean absolute difference to 0.99% and 3.02 mg/dl, respectively. Additionally, the root mean square error and mean absolute relative difference are calculated as 4.81 mg/dl and 1.61%, respectively. The SEG analysis reveals that the proposed method falls within the 100% acceptable zone. Conclusion: The 950 nm non-invasive method for glucose detection has great potential for further development and practical application in the field of glucose monitoring, as it can be further enhanced as a portable device. From a statistical point of view, the collected samples are limited, and more samples need to be collected. © 2023, The Author(s), under exclusive licence to The Brazilian Society of Biomedical Engineering.