Geographical origin discriminant analysis of Chia seeds (Salvia hispanica L.) using hyperspectral imaging

Abstract In this study, hyperspectral imaging was used to study chia seeds grown in Argentina, Paraguay, and Bolivia to determine their origin. Multiplicative scatter correction was the most effective data preprocessing method, with a discrimination prediction accuracy of 0.9111. Discriminative wavelengths obtained from beta coefficients were 912, 1270, 1405, 1611, and 1700 nm, associated O-H bonds in water and C-H bonds of fat and fatty acids). Crude lipid content and fatty acid composition differed significantly by origin. Palmitic, linoleic and α-linolenic acid values were predicted with R2 values of 0.9735, 0.9729 and 0.9526, respectively, with a partial least squares regression model. Thus, the origins of chia seeds can be used to discriminate using hyperspectral imaging analysis.