Real-time monitoring of fat crystallization using pulsed acoustic spectroscopy and supervised machine learning

Opportunities and challenges of ultrasonic diagnostic techniques for plant-based food monitoring: principle, machine system, and application strategies

Plant-based food consumption has increased substantially owing to its positive effects on human and global health. However, ensuring the quality and safety of plant-based foods remains a challenge. Diagnostic ultrasonic technology is widely used for rapid and nondestructive determination owing to its ability to penetrate optically opaque materials, strong directivity, rapid detection capabilities, low equipment costs, and ease of operation. This review provides a comprehensive understanding of diagnostic ultrasonic technology by summarizing the principles of food characterization, factors that influence detection accuracy and methods to mitigate their impact, composition of ultrasonic machine systems, and application of diagnostic ultrasound for monitoring plant-based foods. The detection principle of ultrasonic technology is based on empirical equations that establish a relationship between the ultrasonic and physicochemical indicators of food. To improve the detection accuracy, a compensation mechanism for the temperature and pressure should be established, measurement distances should be set in the far-field region, and liquid samples should be degassed. Furthermore, the sample platform design and the choice of detection mode depend on the nature of the food. Combining ultrasonic technology with machine learning techniques presents promising prospects for real-time process monitoring in the food and beverage industries.

In situ single-droplet analysis of emulsified fat using confocal Raman microscopy: insights into crystal network formation within spatial resolution

Fat crystallization is one of the predominant factors influencing the structure and properties of fat-containing emulsions. In the present study, the role of emulsifiers on fat crystallization dynamics within droplet multiphase systems was evaluated via single-droplet analysis, taking advantage of the non-destructive properties of confocal Raman microscopy. Palm oil droplets dispersed in water were used as a model system, due to palm oil's well-known crystallization properties. Emulsion droplets of the same size were generated using two different emulsifiers (Whey Protein Isolate and Tween 60), at various concentrations. Fast and slow cooling treatments were applied to affect fat crystallisation and network formation as well as droplet morphology, and crystallization dynamics. Raman imaging analysis demonstrated that the chemical structure and concentration of the emulsifier significantly influenced both crystal nucleation within the droplets, as well as the spatial distribution and morphology of the fat crystal network. Additionally, analysis of the spectra of the crystallized phase provided essential information regarding the impact of the emulsifiers on the microstructure, degree of structural order, and structural arrangements of the fat crystal networks. Furthermore, by performing single droplet analysis during cooling it was possible to observe shape distortions in Tween 60 stabilized droplets, as a consequence of the formation of a three-dimensional network of fat crystals that strongly interacted with the interface. On the other hand, the droplets retained their shape when whey proteins were absorbed at the interface. Confocal Raman microscopy, in combination with polarized light microscopy, is, therefore, a well-suited tool for in situ, single-droplet analysis of emulsified oil systems, providing essential information about emulsified fat crystallization dynamics, contributing to better understanding and designing products with enhanced structure and function.