Authentification of Cut Pieces of Chicken Meat by Near Infrared Spectroscopy

• Fumière, O. , Sinnaeve, G. , François, E. & Dardenne, P. (1999). Authentification of Cut Pieces of Chicken Meat by Near Infrared Spectroscopy. Poster in: 9th International Conference on Near Infrared Spectroscopy (ICNIRS), Verona - Italy, 13-18/06/1999.

Type	Poster
Year	1999
Title	Authentification of Cut Pieces of Chicken Meat by Near Infrared Spectroscopy
Event name	9th International Conference on Near Infrared Spectroscopy (ICNIRS)
Event location	Verona - Italy
Label	U15-0804
Recnumber	804
Event date	13-18/06/1999
Endnote Keywords	Chicken|NIR|Authentication|
Abstract	As samples leave a specific fingerprint in the NIR region of the electromagnetic spectrum, NIRS which is a quick, non destructive and cheap technique, is often used to classify the samples using the most recent chemometric techniques (e.g. PCA, factorial discriminant analysis and ANN). As far as meat and meat products are concerned, spectroscopic techniques are used for quantitative determination of major constituents, such as moisture and fat contents (1, 2), for estimating organoleptic quality (3), for objective classification of pig carcasses (4) and for detection of fraudulence (5). The aim of this work is to build statistical models being able to discriminate slow-growing chicken strains meat's spectra from those of " industrial " chicken strains. If the results are positive, the technique could be integrated in an analytical system of surveillance of certified meat products. Chicken meat spectra were obtained with two spectrometers : NIRSystems 6500 (NIRSystems Inc., Silver Spring, MD, USA) and Perten DA 7000 (Perten Instruments Inc., Chatham, IL, USA). All manipulations and processing of the spectra were carried out with the software ISI-NIRS 3 ver. 4.0 ( Infrasoft International, Port Matilda, PA, USA). In this case (discrimination between " slow-growing chicken strains " and " industrial chicken strains "), we used the " Discriminate groups " option, based on partial least square regression (PLS2). Statistical models were made with 2/3 of the samples randomly selected. The last third was used to validate the models. The discriminant models developed on seventy-four chickens show promising performances to distinguish slow-growing strains from industrial strains. Between 70 and 100% are well classified on this basis. The analytical performances of the discriminant models obtained with the NIRSystems 6500 are much better than those obtained with the Perten DA 7000. These results show that NIR spectrometry is a good technique for chicken meat authentication. The authors would like to thank OSTC (Belgian Federal Office for Scientific, Technical and Cultural Affairs) for their financial support.
Fichier
Caption	804-fumiere-1999.ppt
Authors	Fumière, O., Sinnaeve, G., François, E., Dardenne, P.