EVALUATION OF THE QUALITY AND CLASSIFICATION OF PARBOILED RICE USING NEAR-INFRARED SPECTROSCOPY AND MULTIVARIATE STATISTICAL ANALYSES

Abstract

Physical classification is the official standard method for determining grain quality for commercialization. However, it is a time-consuming, subjective operation, susceptible to errors, and requires skilled labor. Optical methods of indirect measurement emerge as a promising evaluation alternative, offering economic advantages, standardization in the assessment of grain nutritional quality, and greater accuracy. Therefore, this study aimed to evaluate the use of near-infrared (NIR) spectroscopy and multivariate statistical analyses to determine the physicochemical quality of parboiled rice grains. Parboiled rice samples were classified according to the Technical Regulation for Rice (Type 1 to Type 5 and Off-Type). Each type was analyzed by NIR to determine the proximate composition (crude protein, moisture, lipids, crude fiber, ash, and starch). The data obtained were subjected to analysis of variance, Tukey’s test, Pearson correlation, and principal component analysis. Regarding starch, the main constituent of rice grains, Types 1 and 2 had the highest concentrations (70.11% and 70.16%, respectively), while the lowest concentrations (66.52% and 66.73%) were found in Types 3 and 5, respectively. The results indicated that NIR, combined with multivariate statistical analyses, can be an efficient alternative for characterizing the physicochemical quality of parboiled rice, highlighting clear patterns, especially in starch and fiber content.