Authentication of raw and cooked freeze-dried rainbow trout (Oncorhynchus mykiss) by means of near infrared spectroscopy and data fusion

This study investigated the possibility of using near infrared spectroscopy (NIRS) for the authentication of raw and cooked freeze-dried rainbow trout (Oncorhynchus mykiss) fillets. Latent variable models applied on the spectral data were developed and used to estimate proximate composition, fatty acid profile, fillet yield and cooking loss, and to classify the available dataset by the rearing farm and genetic strain of each sample. Results showed that NIR spectra can be used both to accurately estimate several chemical properties and to classify samples by rearing farm. In order to classify samples by genetic strain instead, a data fusion approach in which color and mechanical information were combined with spectral data was used. No major differences were observed between the results obtained from raw freeze-dried fillets and those obtained from cooked freeze-dried fillets, except for the estimation of certain chemical constituents of interest such as C22:6 n−3 and polyunsaturated fatty acid content, both of which were better estimated from cooked freeze-dried fillets.