Cookie composition analysis by Fourier transform near infrared spectroscopy coupled to chemometric analysis

The consumption of cookies is ever growing and during the COVID-19 pandemic reached record consumption values and it is imperative to guarantee the quality and safety of the products. Fourier transform near infrared (FT-NIR) spectroscopy, combined with chemometric techniques, provides a promising solution in that regard, due to its speed and simple sample preparation. The objective of this study was to investigate the possibilities of using FT-NIR to predict lipids, carbohydrates, fibers, proteins, salt and energy contents, as well as to identify cookies type and main cereals present in a batch of 120 commercially acquired samples. The prediction models were performed using ordinary least squares (OLS), partial least squares (PLS), and PLS based classification models including discriminant analysis (PLS-DA), k-nearest neighbors (PLS-kNN) and naive Bayes (PLS-NB). The best prediction models allowed for good accuracies, with correlation coefficients higher than 0.9 for all studied nutritional parameters. PLS-kNN methodology was able to identify all 5 main cereals (wheat, integral wheat, oat, corn and rice) as well as the 14 types of cookies based on the nutritional contents. The developed methods were able to accurately identify the cookies type and composition, confirming the proposed methodology as a fast, reliable, environmentally friendly and non-destructive alternative to standard analytical methods.