Quick vacuum drying of liquid samples prior to ATR-FTIR spectral collection improves the quantitative prediction: a case study of milk adulteration

Possible Alternatives: Identifying and Quantifying Adulteration in Buffalo, Goat, and Camel Milk Using Mid-Infrared Spectroscopy Combined with Modern Statistical Machine Learning Methods

Adulteration of higher priced milks with cheaper ones to obtain extra profit can adversely affect consumer health and the market. In this study, pure buffalo milk (BM), goat milk (GM), camel milk (CM), and their mixtures with 5-50% (vol/vol) cow milk or water were used. Mid-infrared spectroscopy (MIRS) combined with modern statistical machine learning was used for the discrimination and quantification of cow milk or water adulteration in BM, GM, and CM. Compared to partial least squares (PLS), modern statistical machine learning-especially support vector machines (SVM), projection pursuit regression (PPR), and Bayesian regularized neural networks (BRNN)-exhibited superior performance for the detection of adulteration. The best prediction models for the different predictive traits are as follows: The binary classification models developed by SVM resulted in differentiation of CM-cow milk, and GM/CM-water mixtures. PLS resulted in differentiation of BM/GM-cow milk and BM-water mixtures. All of the above models have 100% classification accuracy. SVM was used to develop multi-classification models for identifying the high and low proportions of cow milk in BM, GM, and CM, as well as the high and low proportions of water adulteration in BM and GM, with correct classification rates of 94%, 100%, 100%, 99%, and 100%, respectively. In addition, a PLS-based model was developed for identifying the high and low proportions of water adulteration in CM, with correct classification rates of 100%. A regression model for quantifying cow milk in BM was developed using PCA + BRNN, with RMSEV = 5.42%, and RV2 = 0.88. A regression model for quantifying water adulteration in BM was developed using PCA + PPR, with RMSEV = 1.70%, and RV2 = 0.99. Modern statistical machine learning improved the accuracy of MIRS in predicting BM, GM, and CM adulteration more effectively than PLS.

FTIR-based prediction of collagen content in hydrolyzed protein samples

Fourier transform infrared spectroscopy (FTIR) is a powerful analytical tool that has been used for protein and peptide characterization for decades. In the present study, the objective was to investigate if FTIR can be used to predict collagen content in hydrolyzed protein samples. All samples were obtained from enzymatic protein hydrolysis (EPH) of poultry by-products providing a span in collagen content from 0.3% to 37.9% (dry weight), and the FTIR analysis was performed using dry film FTIR. Since nonlinear effects were revealed by calibration using standard partial least squares (PLS) regression, Hierarchical Cluster-based PLS (HC-PLS) calibration models were constructed. The HC-PLS model provided a low prediction error when validated using an independent test set (RMSE = 3.3% collagen), while validation using real industrial samples also showed satisfying results (RMSE = 3.2%). The results corresponded well with previously published FTIR-based studies of collagen, and characteristic spectral features for collagen were well identified in the regression models. Covariance between collagen content and other EPH related processing parameters could also be ruled out in the regression models. To the authors' knowledge, this is the first time that collagen content has been systematically studied in solutions of hydrolysed proteins using FTIR. This is also one of few examples where FTIR is successfully used to quantify protein composition. The dry-film FTIR approach presented in the study is expected to be an important tool in the growing industrial segment that is based on sustainable utilization of collagen-rich biomass.

Discrimination of milk species using Raman spectroscopy coupled with partial least squares discriminant analysis in raw and pasteurized milk

BACKGROUND

Heat treatment is the most common practice for the microbiological safety of milk; hence, determination of the heat treatment of milk is essential. Also, mislabeling or adulteration of expensive milk samples, like ewe or goat milk, with cow's milk is a growing problem in the dairy market. Thus, the determination of the authenticity of milk samples has crucial importance for both producers and consumers. The aim of this study was to discriminate milk samples using Raman spectroscopy with partial least squares discriminant analysis (PLS-DA), first with regard to whether the milk was heat-treated or not, and second with regard to species (cow, goat, ewe, mixture (adulterated)) in both raw and pasteurized milk.

RESULTS

First, discrimination of milk samples as raw or pasteurized was achieved using PLS-DA. Both in calibration and prediction models, high sensitivity and specificity values were obtained for raw and pasteurized milk samples. Second, the proposed method also discriminated milk samples according to their species (cow, goat, ewe, and mixture) for both raw and pasteurized milk. In both calibration and prediction models, the sensitivity and specificity values were above 0.857 and 0.897 respectively. Also, the accuracy values were above 0.915. The results obtained denote satisfactory accurate classification of the samples.

CONCLUSION

The results suggest that Raman spectroscopy coupled with PLS-DA can be successfully used to discriminate milk samples according to heat treatment (raw/pasteurized) and their species within 20 s per sample. It was seen that Raman spectra provide valuable information to be used especially for discrimination of milk samples according to their origin. © 2020 Society of Chemical Industry.