A Review on MIR, NIR, Fluorescence and Raman Spectroscopy Combined with Chemometric Modeling to Predict the Functional Properties of Raw Bovine Milk

Spectroscopic methods, such as Mid-Infrared (MIR), Near-Infrared (NIR), fluorescence and Raman spectroscopy are rapid, inexpensive and nondestructive. Traditionally, mainly MIR and NIR spectroscopy have been employed to predict the compositional properties of milk. However, measurement of the key functional properties of milk is of high industry relevance. In this review, studies on the use of spectroscopic techniques for predicting milk functional properties are compared and reported models are outlined. The challenges of employing spectroscopy in functionality applications are discussed. For pH and curd yield, some of the MIR models display a robust prediction performance. With further model validation, calibrations for these properties could potentially be added to existing MIR instruments in the industry. Despite fluorescence and NIR spectroscopy being used for many dairy applications, their use for milk functionality is limited currently. As Raman spectroscopy is sensitive to the components of raw milk, it has potential for predicting milk functional properties.