Application of Infrared Spectroscopy Technique and Chemometrics for Measurement of Components in Rice after Radiation

摘要

The aim of this study was to investigate the potential for quantitative assessment of amylose and protein content in rice after gamma irradiation using infrared spectroscopy and chemometrics. Rice was treated with eight different radiation doses (250, 500, 750, 1000, 1500, 2000, 2500, and 3000 Gy) and compared to untreated rice (i.e., 0 Gy). Near-infrared (NIR; 1100-2500 nm) and mid-infrared (MIR; 400-4000 cm-1) spectra of the rice were compared to determine which one produced the best prediction of components for irradiated rice. Least-squares support vector machine (LS-SVM) was applied to construct calibration models for component analysis of amylose and protein individually. The optimal results built by LS-SVM were obtained when the rp and RMSEP values were 0.8514 and 0.1519, respectively, for prediction of amylose in the NIR region and 0.8824 and 0.2012, respectively, for prediction of protein in the MIR region. Chemometrics based on LS-SVM are better than that of a back-propagation artificial neural network (BP-ANN). This work demonstrates the potential of infrared reflectance spectroscopy using NIR and MIR for more efficient analysis of components in irradiated rice