FOOD ANALYSIS · Development of novel nondestructive techniques to analyze food quality

Final Activity Report Summary - FOOD ANALYSIS (Development of novel nondestructive techniques to analyze food quality)

There is a continuously growing demand for high-speed systems able to measure quality and safety characteristics of food products, e.g. defects, composition, cheese functional properties or meat tenderness. These systems are supposed to replace labour-intensive and time-consuming analytical techniques and visual as well as manual inspection yet practiced in the food industry at present. Recently developed instruments including spectrometers and imaging devices allow measuring quality attributes of foods in a rapid and non-destructive manner. Moreover, these instruments, combined with sophisticated data processing tools, allow for collecting large amount of experimental data and extracting useful information related to food properties and quality. The project was dedicated to the examination of the bovine meat composition (myofiber, collagen, and fat) and, using different imaging techniques, such as fluorescence, nuclear magnetic resonance, histology and physicochemical techniques, demonstrated a promising potential of fluorescence imaging combined with multivariate statistical tools in the study of beef meat. Partial least square models that were developed indicated relations between extracted image features and mechanical properties and some chemical components of the studied muscles. Based on a system called 'multiway imager' which was previously built by 'École Nationale d'ingénieurs des techniques des industries agricoles et alimentaires' (ENITIAA) in Nantes, France, we tested its potential on meat for characterising the spatial organisation of tissues, namely fat, myofibre and collagen, at the surface of meat samples and on different blue cheeses for characterising the spatial organisation of the blue veins. This system was able to record multivariate images under more than 20 independent spectral conditions. The available channels (plans) of the image were obtained from ultraviolet (UV), visible or near infrared illumination conditions. The resulting images were 'cubes' with two spatial dimensions and one spectral dimension. The results of these studies demonstrated a promising potential of multi-way imaging combined with multivariate statistical tools to discriminate: 1. different muscle types in relation with the spatial organisation of collagen network and myofibrils; and 2. blue cheeses as a function of brand and manufacturer. The results of the project on the study of the properties of experimental hard, semi-hard, and soft cheeses showed the ability of the spectral, both fluorescence and mid-infrared, and physicochemical data recorded on milks and whey to discriminate soft, semi-hard, and hard cheeses. In addition, the spectral bands allowing for cheeses' discrimination were identified and correlated with measured physicochemical parameters using chemometric methods such as common components and specific weights analysis. Finally, literature search was conducted and showed that spectroscopic techniques could be suitable for the quantification of essential nutrients in foods. This might be an interesting field for the development of rapid techniques for the food industry.

Data: CORDIS, © European Union