Application of multispectral imaging (MSI) to food and feed sampling and analysis

Last updated:
21 July 2017
Multispectral imaging (MSI) represents an innovative and non-invasive technique combining both imaging and spectral technologies. It uses multiple wavelengths from the UV through to the near infra-red spectrum, permitting fast and accurate determination of surface colour, texture and possible chemical composition.
Study duration: January 2016 to November 2016
Project code: FS301017
Contractor: LGC
Contact:

For any enquiries relating to this project please contact SERD@foodstandards.gsi.gov.uk

 

Background

Effective sampling and analysis is an important aspect of official food and feed controls. Traditionally, there are a number of physio-chemical and biological methods available for monitoring food and feed, but many of these techniques are expensive, invasive, technically demanding, and are often associated with long turn-around times.

Compared to traditional molecular biology approaches utilising DNA extraction followed by PCR-based analyses, MSI can simplify and reduce the time and costs associated with sample analysis. Other advantages of MSI include that it is rapid and non-destructive, a true multi-analyte non-targeted approach, non-invasive, well suited to automation, has limited requirement for operator/analyst training and lower consumable costs.

Research Approach

The project mainly aims to provide actionable evidence to the practical implementation of MSI to topical authenticity, quality and safety issues associated with food and feed sampling and analysis in the UK.

Key objectives include the identification and prioritisation of core UK authenticity, quality and safety issues associated with the sampling and analysis of food and feed through discussions with stakeholders, sourcing and developing a range of appropriate test materials, and evaluating the performance capabilities of MSI to characterise the panel of representative test samples. MSI offers the potential to significantly reduce both the timeframes currently required for testing samples and the incumbent cost burden to the UK Government and commercial testing laboratories. This technological development helps reinforce the overarching aims of the FSA Strategic Plan 2015-2020, lending itself well to supporting the initiative of making food safe to eat, and protecting consumers’ interests in relation to food labelling and quality.

The project will be carried out using the following approach:

  • Identify a priority list of food and feed sample testing scenarios, regarding authenticity, quality and safety issues associated with sampling and analysis of food and feed in the UK. This will be achieved through an in-depth literature review and UK stakeholder engagement.
  • Source and provide relevant food and feed samples that are representative of the testing scenarios identified as part of the priority list mentioned above.
  • Initial study: Evaluate the capability of MSI to assess the top three testing scenarios in the priority list of food and feed sample issues, and provide interim report to FSA.
  • Full study: Evaluate the capability of MSI to assess remaining testing scenarios in the priority list of food and feed samples, representative of different levels of adulteration, quality and safety issues.
  • Provide a final report on the demonstrable application of MSI inclusive of recommendations on how UK stakeholders can implement the technology.

Results

This Proof-of-Principle project provides evidence to suggest that MSI could be used as a rapid, automated and cost-effective screening tool to help augment pre-existing approaches for food authenticity, adulteration and quality testing, by a range of UK laboratories.

The applicability of MSI is dependent on the spectral and physical properties of the component materials, and the image analyses methodologies applied. The test scenarios explored covered a wide variety of challenging test materials with differing physiochemical characteristics. The results clearly show that the MSI methodologies applied were capable of detecting and differentiating between the majority of test components and provided preliminary quantitative data.

A summary of the estimated limits of sensitivity for the samples assessed in this project is presented in the following table:

Sampling Scenario Sample (adulterant)

Limit of detection (LOD) (w/w)
Oregano (olive leaves) ≤ 0.5%
Rice (Plastic rice) ≤ 1%
Rice (gravel) ≤ 1%
Beef (pork) ≤ 5%
Beef (offal) ≤ 10%
Ground cumin (ground almond shell) ≤ 5%
Ground cumin (ground mahaleb) ≤ 5%
Ground almond (ground peanut) ≤ 5%
Wheat flour (peanut flour) ≤ 10%

MSI has demonstrated good potential to be used as a screening method for a range of sampling scenarios (detailed in the above table), and has additional functionality when an adulterant or contaminant is present above a certain threshold level. Clearly MSI will not be applicable as a screening approach for every conceivable sampling situation (food authenticity, adulteration, quality and safety testing), and the added advantage of the MSI’s quantitative capability would benefit from further characterisation. Full method validation on selected sampling scenarios and evaluation of the quantitative accuracy of MSI would facilitate an objective judgement of its fitness for purpose and applicability on a sample case by case basis. This initial feasibility study demonstrated the utilisation of MSI across a range of different sampling scenarios, inclusive of its utilisation for testing for contaminants, but additional work is recommended in order to fully validate it fitness for purpose on a sample by sample basis.