Selectivity-enhanced glucose measurement in multicomponent aqueous solution by broadband dielectric spectroscopy

We demonstrate the detection of physiological-range glucose in a multicomponent aqueous solution through multivariate analysis of broadband dielectric spectra from 500 MHz to 50 GHz. To enhance the selectivity to glucose, we applied spectral preprocessing on dielectric spectra to extract the feature values of glucose and bovine serum albumin (BSA). Using the regression models derived from different concentrations of glucose and BSA, the analysis was carried out on the solutions with the physiological range of both components. The prediction error of glucose concentration was estimated at less than 73 mg/dL even in various concentrations of BSA. This technique is easily implemented with a microwave blood glucose sensor or in other biomedical applications that essentially require multicomponent analysis.