Development of a non-invasive blood glucose sensor using short-wavelength near-infrared spectroscopy and its application to glycemic index determination

Abstract

Glycemic index (GI) determination requires dozens of capillary blood samplings by fingertip puncture, resulting in patient discomfort. Therefore, a new method without painful capillary blood samplings is highly desirable. In this study, a non-invasive blood glucose sensor using short-wavelength near-infrared (NIR) spectroscopy was developed and its application to glycemic index determination was examined. A non-invasive blood glucose sensor was developed based on the existing spectrophotometer with fiber optics in interactance mode. In order to avoid spectral variation, measuring area, contact pressure and sample temperature were kept constant. NIR spectra in the short-wavelength region from 700 nm to 1 050 nm were measured on the palm of the hand during standard food tolerance tests as part of GI determination. Partial least squares (PLS) regression was employed to derive an individual calibration model for determining blood glucose content. The standard error of cross-validation (SECV) for the developed calibration model was 9.1 mg/dL. The GIs of three test foods, white rice, boiled fish paste and yogurt, calculated using the blood glucose values measured by the developed calibration model, were 70 (calculated using actual values = 80), 57 (49), and 45 (38), respectively. The non-invasive blood glucose sensor using short-wavelength NIR spectros-copy shows promise for its applicability to GI determination.