34136

Application of near-infrared hyperspectral imaging for assessment of oxidative degradation in Atlantic salmon (Salmo salar) fillets during cold storage

Favoritar este trabalho

This study investigated the potential of applying hyperspectral imaging technology (900-1,700 nm) to determine the thiobarbituric acid (TBA) value and pH for monitoring and evaluation of lipid oxidation in Atlantic salmon fillets during cold storage. In order to observe the changes of lipid oxidation from sample to sample even from spot to spot within the same sample, it is quite important to visualise the spatial distribution of TBA values. By integrating both spectroscopic and imaging techniques in one system, hyperspectral imaging is able to generate a spatial map of spectral variation. Therefore, it is a powerful extension of traditional analytical techniques to study the distribution of different quality attributes in one sample, resulting in many successful applications in the quality control of many food products. In this study, a total of 150 fresh Atlantic salmon (Salmon Salar) fillets were prepared and randomly divided into five groups subjected to cold storage for 0,3,6,9 and 12 days in a refrigerator. After scanned by hyperspectral imaging machine, the reference values of pH and TBA were immediately determined. The pH value was evaluated by a benchtop pH meter (Orion 520Aplus, US), after which TBA value was determined according to the procedure described by Vyncke (1970) with some modifications. The PLSR calibration models were built with full spectral region and showed good performance for simultaneously predicting TBA value and pH with root-mean-square errors of prediction (RMSEP) of 2.369 µmol MDA/kg fish and 0.0709, respectively. Two simplified MLR models for pH and TBA value were built and compared using the selected 10 and 18 most important wavelengths. The optimised MLR model for pH showed a little better performance with correlation coefficients of 0.830 and RMSEP of 0.0483. Meanwhile, the optimised MLR model for TBA value yielded satisfactory results with correlation coefficients of 0.921and RMSEP of 2.081 µmol MDA/kg flesh, which was used to visualise the TBA values distribution in fish fillets. The distribution maps showed that there was a general trend of increment in overall TBA values from storage day 0 to 12. In addition, it was also observed that the distribution density of TBA value was non-uniform and asymmetric. For instance, the ventral section displaying more intensified red colour were more inclined to lipid oxidation compared with the dorsal part of the fillet. In brief, this visualisation of TBA value distribution has successfully provided dynamic changes of lipid oxidation and freshness loss of salmon fillet during cold storage. The results from this work confirmed that near-infrared hyperspectral imaging technique is suitable for the determination of TBA values and pH to monitor lipid oxidation and evaluate fish freshness.