An improved integration sensor of non-invasive blood glucose

This work is intended to design an integration sensor for non-invasive blood glucose detection (ISNBGD) with low power consumption, fast detection speed and high accuracy. Based on the modified conservation of energy metabolism theory, the blood glucose level (BGL) is acquired by a multi-linear regression equation of a rate of human metabolism, blood flow, oxygen saturation of hemoglobin, and pulse frequency. These physiological parameters are obtained by ISNBGD with a size of 60mm × 30mm × 85.5mm, which is integrated with temperature sensors, humidity sensors, radiation sensors, and dual wavelength photo-detectors with 660nm and 905nm. 30 healthy volunteers are involved in the preliminary clinical experiment. These blood glucose concentrations ranging from 3.5 to 6.5mmol/L are compared with automatic biochemical analyzer (ABA) as the gold standard. The correlation coefficient of BGL is up to 0.8621 between the ISNBGD and ABA, and the ISNBGD has strong correlation and good coherence. Moreover, multi-physiological parameters can be measured by the ISNBGD including body temperature, local metabolic rate, blood flow, oxygen saturation, and heart rate simultaneously.