Prediction of bioactive compounds in barley by near-infrared reflectance spectroscopy (NIRS)

Barley grains contain a variable amount of biologically active compounds such as non-starch polysaccharides and phenol compounds. These are important in nutrition due to their significant health benefits technological role food. We developed predictive models for ?-glucans (BG), arabinoxylans (AX), bound phenols (BP), free (FP), anthocyanins (AN) based on near-infrared spectroscopy (NIRS) using two different NIRS instruments with spectral range steps. Regressions modified partial least squares (MPLS) several combinations scattering correction derivative treatments were tested. The optimal calibration generated high coefficients determination BG BP, but not AN content. instrument the highest resolution only gave better results prediction models, addition visible did prove be ostensibly advantageous any study, even case analysis.