Profile of lactic acid bacteria (MALDI-TOF-MS) and physico-chemical and microbiological characteristics of the raw milk and fresh artisanal cheese from Serra Geral, Minas Gerais, Brazil.

Artisanal cheese from Serra Geral, Minas Gerais, Brazil, stands out for its cultural asset and socio-economic relevance. However, standards of identity and quality and the peculiar terroir associated with the edaphoclimatic conditions have not been established. Therefore, the production flow diagram and the physico-chemical and microbiological quality of the raw milk, pingo (natural starter culture), production benches, water and fresh cheese were investigated for the first time. In addition, lactic acid bacteria (LAB) from cheese and its production environment were identified by MALDI-TOF. For that, 12 cheese making facilities were selected. The raw milk and pingo showed adequate physico-chemical characteristics for cheesemaking; however, high microbial counts were found. In the water, total and thermotolerant coliforms were also identified. The fresh cheeses were classified as 'high moisture and fat' and 'soft mass'. Most physico-chemical parameters were satisfactory; however, there were high counts of total coliforms, Staphylococcus spp. and coagulase-positive staphylococci. There were high counts of LAB in the raw milk, pingo, bench surface and fresh cheese. A total of 84 microbial biotypes from MRS agar were isolated. Lactococcus lactis was the predominant LAB, followed by Lactococcus garvieae. Leuconostoc mesenteroides (benches), Leuconostoc pseudomesenteroides (fresh cheese), and Enterococcus faecium (pingo) were identified sporadically. These results indicate the risks to public health associated with the consumption of the fresh cheese, and measures to improve its safety are needed.

Determination of the lactose content in low-lactose milk using Fourier-transform infrared spectroscopy (FTIR) and convolutional neural network.

Demand for low lactose milk and milk products has been increasing worldwide due to the high number of people with lactose intolerance. These low lactose dairy foods require fast, low-cost and efficient methods for sugar quantification. However, available methods do not meet all these requirements. In this work, we propose the association of FTIR (Fourier Transform Infrared) spectroscopy with artificial intelligence to identify and quantify residual lactose and other sugars in milk. Convolutional neural networks (CNN) were built from the infrared spectra without preprocessing the data using hyperparameter adjustment and saliency map. For the quantitative prediction of the sugars in milk, a regression model was proposed, while for the qualitative assessment, a classification model was used. Raw, pasteurized and ultra-high temperature (UHT) milk was added with lactose, glucose, and galactose in six concentrations (0.1-7.0 mg mL-1) and, in total, 432 samples were submitted to convolutional neural network. Accuracy, precision, sensitivity, specificity, root mean square error, mean square error, mean absolute error, and coefficient of determination (R2) were used as evaluation parameters. The algorithms indicated a predictive capacity (accuracy) above 95% for classification, and R2 of 81%, 86%, and 92% for respectively, lactose, glucose, and galactose quantification. Our results showed that the association of FTIR spectra with artificial intelligence tools, such as CNN, is an efficient, quick, and low-cost methodology for quantifying lactose and other sugars in milk.

A machine learning proposal method to detect milk tainted with cheese whey.

Cheese whey addition to milk is a type of fraud with high prevalence and severe economic effects, resulting in low yield for dairy products, nutritional reduction of milk and milk-derived products, and even some safety concerns. Nevertheless, methods to detect fraudulent addition of cheese whey to milk are expensive and time consuming, and are thus ineffective as screening methods. The Fourier-transform infrared (FTIR) spectroscopy technique is a promising alternative to identify this type of fraud because a large number of data are generated, and useful information might be extracted to be used by machine learning models. The objective of this work was to evaluate the use of FTIR with machine learning methods, such as classification tree and multilayer perceptron neural networks to detect the addition of cheese whey to milk. A total of 520 samples of raw milk were added with cheese whey in concentrations of 1, 2, 5, 10, 15, 20, 25, and 30%; and 65 samples were used as control. The samples were stored at 7, 20, and 30°C for 0, 24, 48, 72, and 168 h, and analyzed using FTIR equipment. Complementary results of 520 samples of authentic raw milk were used. Selected components (fat, protein, casein, lactose, total solids, and solids nonfat) and freezing point (°C) were predicted using FTIR and then used as input features for the machine learning algorithms. Performance metrics included accuracy as high as 96.2% for CART (classification and regression trees) and 97.8% for multilayer perceptron neural networks, with precision, sensitivity, and specificity above 95% for both methods. The use of milk composition and freezing point predicted using FTIR, associated with machine learning techniques, was highly efficient to differentiate authentic milk from samples added with cheese whey. The results indicate that this is a potential method to be used as a high-performance screening process to detected milk adulterated with cheese whey in milk quality laboratories.

On the utilization of deep and ensemble learning to detect milk adulteration.

BACKGROUND: Fraudulent milk adulteration is a dangerous practice in the dairy industry that is harmful to consumers since milk is one of the most consumed food products. Milk quality can be assessed by Fourier Transformed Infrared Spectroscopy (FTIR), a simple and fast method for obtaining its compositional information. The spectral data produced by this technique can be explored using machine learning methods, such as neural networks and decision trees, in order to create models that represent the characteristics of pure and adulterated milk samples. RESULTS: Thousands of milk samples were collected, some of them were manually adulterated with five different substances and subjected to infrared spectroscopy. This technique produced spectral data from the milk samples composition, which were used for training different machine learning algorithms, such as deep and ensemble decision tree learners. The proposed method is used to predict the presence of adulterants in a binary classification problem and also the specific assessment of which of five adulterants was found through multiclass classification. In deep learning, we propose a Convolutional Neural Network architecture that needs no preprocessing on spectral data. Classifiers evaluated show promising results, with classification accuracies up to 98.76%, outperforming commonly used classical learning methods. CONCLUSIONS: The proposed methodology uses machine learning techniques on milk spectral data. It is able to predict common adulterations that occur in the dairy industry. Both deep and ensemble tree learners were evaluated considering binary and multiclass classifications and the results were compared. The proposed neural network architecture is able to outperform the composition recognition made by the FTIR equipment and by commonly used methods in the dairy industry.

Survey of pyrethroid, macrocyclic lactone and antibacterial residues in bulk milk tank from Minas Gerais State, Brazil / Resíduos de piretróides, lactonas macrocíclicas e antimicrobianos em amostras de leite de tanque no estado de Minas Gerais

A survey of veterinary drug residues in bulk milk tank from Minas Gerais State, Brazil, was carried out through a broad scope analysis. Here, 132 raw milk samples were collected at 45 dairy farms in Minas Gerais from August 2009 to February 2010, and analyzed for 42 analytes, comprising pyrethroids, macrocyclic lactones and antibacterials, using liquid chromatography coupled with mass spectrometry in tandem mode and gas chromatography with electron capture detection. Within all milk samples, at least one veterinary drug residue was identified in 40 milk samples (30.30%) by confirmatory tests, whereas 16 samples (12.12%) showed the presence of at least two residues. With regard to the Brazilian maximum residue levels, 11 milk samples (8.33%) were non-compliant according to Brazilian Legislation. The veterinary drugs detected in the non-compliant milk samples include penicillin V (one sample), abamectin (one sample) and cypermethrin (nine samples). Furthermore, the antibacterial screening methods failed to identify most of the positive samples that were detected by confirmatory tests, leading to a large discrepancy between the screening and confirmatory antimicrobial tests. Thus, the present study indicated that the veterinary drugs residues still represents a great concern for the milk production chain.(AU)

Avaliou-se a presença de 42 analitos, incluindo piretróides, lactonas macrocíclicas e antimicrobianos em 132 amostras de leite de tanque proveniente de 45 propriedades leiteiras localizadas no Estado de Minas Gerais. Para tal, utilizou-se a cromatografia líquida acoplada a espectrofotometria de massas tandem e cromatografia gasosa com detector com captura de elétrons. Dentre todas as amostras de leite, 40 (30,30%) amostras de leite de tanque apresentaram a presença de pelo menos um analito, enquanto 16 amostras (12,12%) de leite demonstraram a presença de pelo menos dois analitos. Considerando os limites estabelecidos pela legislação brasileira, 11 amostras de leite (8,33%) seriam consideradas como não conforme. Ademais, os testes de triagem para detecção de antimicrobianos no leite não conseguiram identificar a maioria das amostras positivas nos testes confirmatórios, levando a grande discrepância entre estes testes. Desta forma, os resultados do presente estudo indicam que os períodos de descarte do leite, especialmente para piretróides, não foram plenamente respeitados por todos os produtores de leite. Além disto, uma discrepância entre os resultados dos testes confirmatórios e os testes de triagem foi observada.(AU)

Partial budget analysis of prepartum antimicrobial therapy and Escherichia coli J5 vaccination of dairy heifers and their effect on milk production and milk quality parameters / Análise econômica da terapia antimicrobiana no pré-parto e da vacinação com Escherichia coli J5 em novilhas leiteiras e seu efeito sobre a produção e qualidade de leite

This study aimed to determine whether prepartum antimicrobial and/or Escherichia coli J5 vaccination in dairy heifers influence the milk production, milk quality, and estimate their economic benefit. Thus, 33 dairy heifers were enrolled in four groups using a split-splot design. Groups were: (G1) prepartum antimicrobial infusion and vaccination with an E. coli J5 bacterin, (G2) prepartum antimicrobial infusion, (G3) vaccination with an E. coli J5 bacterin, and (G4) control heifers. Composite milk samples for somatic cell count, total bacteria count and milk composition were collected 15 days after calving and every 15 days until the end of the experiment. Bacteriological analysis was carried out at the end of study. The milk production and the incidence of clinical cases of mastitis, as well as the costs associated with them were recorded. The results demonstrate a reduction on clinical mastitis rates by preventive strategies, which implicated in lower volume of discarded milk (0.99, 1.01, 1.04 and 3.98% for G1, G2, G3 and G4, respectively) and higher economic benefit. Thus, in well-managed dairy herds the prevention of heifer mastitis by vaccination or antimicrobial therapy can reduce the amount of antimicrobials needed to treat clinical mastitis cases and the days of discarded milk.

O presente estudo objetivou realizar uma análise econômica do tratamento antimicrobiano no pré-parto e/ou da vacinação com Escherihia coli J5 em novilhas leiteiras, e seu efeito sobre a produção e qualidade de leite. Portanto, utilizou-se o delineamento split-splot em esquema fatorial, no qual 33 novilhas da raça Holandesa foram divididas aleatoriamente em quatro grupos: (G1) antimicroianoterapia no pré-parto e vacinação com E. coli J5, (G2) antimicrobianoterapia no pré-parto, (G3) vacinação com E. coli J5 e (G4) controle. Amostras compostas de leite foram coletadas para contagem de células somáticas, contagem bacteriana total e composição do leite 15 dias após o parto, e a cada 15 dias até o término do experimento. A análise bacteriológica do leite foi realizada ao término do experimento. A produção de leite e a incidência dos casos clínicos de mastite, assim como, os custos associados à antimicrobianoterapia no pré-parto e/ou vacinação com E. coli J5 foram registrados. Os resultados demonstraram redução dos casos clínicos de mastite com a implementação das medidas preventivas resultando no menor volume de leite descartado (0,99, 1,01, 1,04 e 3,98% para os animais dos grupos G1, G2, G3 e G4, respectivemente) e maior benefício econômico. Desta forma, em rebanhos bem manejados, a implementação da antimicrobianoterapia no pré-parto e vacinação com E. coli J5 e novilhas pode reduzir a quantidade de antimicrobianos necessário para o tratamento de casos de mastite clínica durante a lactação, resultando em menor número de dias em que o leite é descartado.

Influence of race and crossbreeding on casein micelles size.

Casein (CN) micelles are colloidal aggregates of protein dispersed in milk, the importance of which in the dairy industry is related to functionality and yield in dairy products. The objective of this work was to investigate the correlation of milk CN micelles diameter from Holstein and Zebu crossbreds with milk composition (protein, fat, lactose, total and nonfat solids and milk urea nitrogen), somatic cell count (SCC), age, lactation stage and production. Average casein micelles diameters of milk samples obtained from 200 cows were measured using photon correlation spectroscopy and multiple regression analysis was used to find relationship between variables. CN micelle diameter, SCC and nonfat solids were different between animals with different Holstein crossbreed ratios, which suggests influence of genetic factors, mammary gland health and milk composition. Overall, results indicate the potential use of CN micelle diameter as a tool to select animals to produce milk more suitable to cheese production.

Homogenization and lipase treatment of milk and resulting methyl ketone generation in blue cheese.

A specific range of methyl ketones contribute to the distinctive flavor of traditional blue cheeses. These ketones are metabolites of lipid metabolism by Penicillium mold added to cheese for this purpose. Two processes, namely, the homogenization of milk fat and the addition of exogenous lipase enzymes, are traditionally applied measures to control the formation of methyl ketones in blue cheese. There exists little scientific validation of the actual effects of these treatments on methyl ketone development. The present study evaluated the effects of milk fat homogenization and lipase treatments on methyl ketone and free fatty acid development using sensory methods and the comparison of selected volatile quantities using gas chromatography. Initial work was conducted using a blue cheese system model; subsequent work was conducted with manufactured blue cheese. In general, there were modest effects of homogenization and lipase treatments on free fatty acid (FFA) and methyl ketone concentrations in blue cheese. Blue cheese treatments involving Penicillium roqueforti lipase with homogenized milk yielded higher FFA and methyl ketone levels, for example, a ∼20-fold increase for hexanoic acid and a 3-fold increase in 2-pentanone.