Monitoring freshness of crayfish (Prokaryophyllus clarkii) through the combination of near-infrared spectroscopy and chemometric method

A method for monitoring the freshness of crayfish by near-infrared spectroscopy combined with chemometrics was proposed in this study. The total volatile basic nitrogen (TVBN), K value and biogenic amine were adopted as the indexes of freshness. The 120 peeled and the 120 whole tail samples were utilized to build the quantitative models by partial least squares (PLS) regression, respectively. The spectral pretreatment and the variable selection methods were adopted to further improve the models. For the models of TVBN and K value, the combination of continuous wavelet transform (CWT) and randomization test (RT) method could improve the predictability effectively. The combination of 1st derivative and RT method could get better results for the optimization of the biogenic amine models. The optimized models were evaluated by 40 peeled and the 40 whole tail samples, and then utilized to predict the three indexes in another 32 peeled tails and 40 whole tail samples for practical application. Reasonable results were obtained for the samples with TVBN levels ranging from 28.01 mg/kg to 284.89 mg/kg, K value levels ranging from 13.04 to 87.36% and biogenic amines ranging from 148.57 mg/kg to 663.71 mg/kg. For the TVBN, K value and biogenic amine in the 32 peeled tail samples, the correlation coefficients between the results calculated by the optimized models and those measured by the standard methods were 0.95, 0.94 and 0.96 respectively. As to the three freshness indexes in 40 whole tail samples, the correlation coefficients were 0.95, 0.93 and 0.94, respectively.