Enhancing quality and safety monitoring of cold chain foods through integrating chemometrics with spectroscopy: concepts and practical implementations

Abstract

Maintaining a robust cold chain to ensure adequate food quality and safety, minimise food losses, and optimise resource and energy utilisation remains a key objective in the food industry. The application of spectroscopic and chemometric techniques for food quality analysis within the cold chain has shown significant potential over traditional methods, owing to their rapid, non-invasive, and reliable evaluation capabilities. This review systematically presents the principles of various chemometric approaches such as principal component analysis (PCA), cluster analysis (CA), k-nearest neighbors (KNN), artificial neural networks (ANNs), and partial least squares (PLS) in combination with spectroscopic technologies, including visible/near-infrared (Vis/NIR), infrared (IR), mid-infrared (MIR), nuclear magnetic resonance (NMR), hyperspectral imaging (HSI), Fourier transform infrared (FTIR), and Raman spectroscopy, for the qualitative and quantitative assessment of food quality. By placing particular emphasis on the application of PCA and PLS in evaluating food quality during the cooling, freezing, and storage stages of the cold chain, several studies demonstrate that this combined approach offers considerable promise for rapid analysis of complex datasets, effective quality monitoring, food authentication, and overall cold chain optimisation. This review uniquely integrates chemometrics and spectroscopy applications across the cold chain, providing a consolidated framework that has not been systematically presented in prior reviews. Furthermore, it identifies key challenges and outlines future research directions aimed at enhancing the reliability, accuracy, and industrial applicability of these techniques across diverse cold chain scenarios.