Abstract
Analytical fingerprinting in combination with chemometric modelling provides a powerful approach in the framework of food quality assurance. This recently emerging approach allows verifying claims regarding a whole range of food quality characteristics that were previously difficult or impossible to determine. These include claims regarding the nutritional composition, product typicality and method of production as well as the geographical origin of food. As such, the fingerprinting approach can for example be used to verify claims of fish being from a wild origin, or of eggs being organic. Since the fingerprinting approach differs from the classical analytical approach for food quality assurance, the existing protocols for validation of classical analytical methods are not directly applicable. To further standardize and harmonize the fingerprinting approach on an international level, an extension to existing validation protocols is required. This chapter provides a first step towards a further standardization by providing a tentative strategy for integral validation of the fingerprinting approach.
Original language | English |
---|---|
Title of host publication | Mathematical and Statistical Methods in Food Science and Technology |
Editors | Gastón Ares |
Publisher | Wiley |
Chapter | 23 |
Pages | 449-470 |
ISBN (Electronic) | 9781118434635 |
ISBN (Print) | 9781118433683 |
DOIs | |
Publication status | Published - 27 Dec 2013 |
Keywords
- Bootstrapping
- Classification model
- Cross-validation
- Decision limit (CCα)
- EC Directive 2002/657/EC, ISO 17025
- External validation
- Method validation
- Performance characteristics
- Repeatability
- Reproducibility
- Ruggedness, Stability
- Selectivity, Detection limit (CCβ)
- Sensitivity
- Specificity