A novel NIR spectroscopic method for rapid analyses of lycopene, total acid, sugar, phenols and antioxidant activity in dehydrated tomato samples

Abstract Near-infrared (NIR) spectroscopy is a well-known, rapid and non-destructive technique suitable for analyses of many different food products. In this work, it was used to develop a novel method for the analysis of dehydrated tomato samples, collected from four different producers. The NIR spectra from such samples were discriminated according to the four production sources, with the use of principal component analysis (PCA). Also, hierarchical cluster analysis (HCA), linear discriminant analysis (LDA) and K-nearest neighbors (KNN) were successfully used for pattern recognition. The prediction rates were 100% for assigning producers to samples. Two multivariate calibration models—partial least squares regression (PLSR) and radial basis function neural networks (RBF-NN), were applied for quantitative analysis. NIRS calibration models were established for the determination of lycopene and total acid, sugar, phenols and antioxidant activity in dehydrated tomatoes. These calibrations were then used for prediction of unknown, dehydrated tomato samples with satisfactory results. The RBF-NN results were better than those obtained from the PLSR models, and the better predictions suggested that the novel NIR spectroscopic method supported by chemometrics, is suitable for the discrimination and prediction of the five quality parameters in the dehydrated tomato products.