We measure what you see - and beyond!

Multispectral Imaging for Food and Feed Analysis

Multispectral Imaging for Food and Feed Analysis

paprika

How can spectral imaging aid food safety analysis?

A recent Food Standards Agency report, conducted by LGC Ltd, used the VideometerLab technology to answer this question. Food fraud, such as adulteration or contamination, is becoming a large-scale problem. Hence, the decision to investigate how to efficiently test food with spectral imaging technology by the British Life Sciences Measurement Organization.

Food adulteration is a common fraudulent practice, in which producers use cheaper or more available generic ingredients, to deceive the customers, or to cut the production cost. In most cases, it is very difficult to distinguish adulterated food from unadulterated food by utilizing traditional sensorial methods.

The focal point of this study was, therefore, to investigate if spectral imaging-based technologies are sufficient to analyze the listed examples of products below to ensure food safety for consumers. As mentioned before, The spectral imaging system used in the study was the VideometerLab, which was chosen because it ”represents cutting-edge technology with intuitive analytical workflows.

Moreover,  the samples used in this report are the most common instances of food fraud around Europe. (FAO Report, 2021). The FSA-supported project established six validated methods including

  • oregano adulteration,
  • presence of offal in meat,
  • beef adulteration with pork,  
  • ground peanut found in ground almond,
  • paprika adulteration,
  • ‘’white’’ fish speciation.

Food adulteration, authenticity, safety & more

During the model development stage, the researchers decided on key areas which to take into consideration for the scope of the study – food-related variables such as suppliers, breed types, or processes. Furthermore, this methodology was also used to validate the results of the project, in order to ensure that the results are reproducible and not limited to certain characteristics.  

The LGC Limited report, Validation of multispectral imaging (MSI) technology for food and feed analysis, explores each of the examples in detail by their methodology, sample processing, analysis as well as validation methods. Below are the key points from the conducted study.

Oregano herb adulteration with olive/myrtle leaves

The oregano used for testing was sourced from various suppliers to ensure data reliability. Furthermore, adulterants – such as olive, myrtle leaves, or sawdust – were mixed with the oregano. The result of the analysis determined that ‘’[Spectral imaging] is capable of detecting adulteration from myrtle with a detection level of <10% myrtle in oregano (w/w).’’ Moreover, the uncertainty of the method was established to be between 1.45 – 4.53%.

Beef meat adulteration with offal

The samples for the analysis came from British supermarkets, local butchers, and organic suppliers. This allowed the study to be more adequate and thorough. The model was developed with known potential adulterants, offal adulterants, and other meat parts.  

The analysis showed that only beef meat inspection can be a validated method, as pork meat exhibited cross-reactivity with other meat materials. The results of meat adulteration with offal show that all test samples were classified correctly.

The sub-study of this section concerns the impact of the freeze-thawing process on meat features. The materials were stored for a minimum of 48 hours at -20OC, subsequently thawed for around 1 hour, and scanned again with the VideometerLab.

You can read more about a similar study and its results that we presented here.

Ground almond contaminated with ground peanut

Materials for samples originated from Spain and the USA. All of them were deshelled and then grounded to the desired consistency for the research. Furthermore, other nuts like cashew, hazelnut, and walnut were also used.  After the validation process, the researchers stated that the performance of the ground peanut classification model exceeded their expectations.

Paprika adulterated with ground almond

Various types of spice were used, including: sweet, smoked paprika, chili, cayenne pepper, as well as different ASTA color values. The ground almond sample was grounded again to the desired consistency for the purpose of the research and then analyzed with VideometerLab.

Using spectral differences, the researchers were able to classify and distinguish contaminated paprika. In consequence, the analysis of these samples ”showed that the methods were […] demonstrated promising performance characteristics suited to initial screening techniques for multiple targets.’’

‘’White’’ fish speciation

Fish for the study were chosen based on the common availability at British ports: fresh cod fillet, haddock fillet, pollack fillet, hake fillet, and whiting fillet. After being trimmed, they were stored at appropriate storage temperatures.

Comparing reflectance spectra for selected materials showed good potential for various fish sample classification, however, it was noticed that the validation part was too large to exhibit a positive conclusion. 

Conclusion

The researchers used the VideometerLab software built-in feature of calculation of area fraction. This allows establishing if the sample matches the specifications of its classification. Thus, the researchers were able to determine a percentage of adulteration in each sample based on the surface area only.  

The LGC report concluded that ‘’the MSI methodologies applied were capable of detecting and differentiation across a range of different samples and test components.‘’ Furthermore, when compared to other competitors, the researchers concluded that ‘’The VideometerLab 4 is a fully integrated desktop portable platform designed to minimize sampling and analytical variability, e.g. standardized analytical workflows.’’

References

Validation of Multispectral Imaging (MSI) technology for food and feed analysis. (2021). The LGC Limited.

Food and Fraud intention, detection, and management. (2021). Food and Agriculture Organization of the United Nations.

 

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: