Abstract
Non-invasive techniques for the analysis of blood glucose are desirable not only for clinical use but also for the determination of the glycaemic index (GI) of foods. Factors that affect the accuracy of non-invasive blood glucose measurements by near infrared (NIR) spectroscopy were identified and examined to establish the optimal conditions for the analysis. Calibration models were obtained by means of partial least squares regression. Several experiments revealed that a good calibration model could be obtained using spectra (700–1100 nm) acquired from the palm of the hand with a large interactance probe fixed to a table provided that spectral acquisition and blood glucose analysis are performed simultaneously. The accuracy of the calibration model developed is clinically acceptable; the coefficient of multiple determination was 0.85 and the standard error of cross-validation was 9.7 mg dL−1. An analysis of a model solution confirmed that NIR spectroscopy in the short-wavelength region 700–1100 nm is capable of measuring glucose concentrations of the order of 0.01% (10 mg dL−1) in blood. The determination of the GI of foods by using the present method will be examined in the near future.
© 2010 IM Publications LLP
PDF Article
More Like This
Cited By
You do not have subscription access to this journal. Cited by links are available to subscribers only. You may subscribe either as an Optica member, or as an authorized user of your institution.
Contact your librarian or system administrator
or
Login to access Optica Member Subscription
Add to BibSonomy