Fusing 1H NMR and Raman experimental data for the improvement of wine recognition models.

The present study proposes the development of new wine recognition models based on Artificial Intelligence (AI) applied to the mid-level data fusion of 1H NMR and Raman data. In this regard, a supervised machine learning method, namely Support Vector Machines (SVMs), was applied for classifying wine samples with respect to the cultivar, vintage, and geographical origin. Because the association between the two data sources generated an input space with a high dimensionality, a feature selection algorithm was employed to identify the most relevant discriminant markers for each wine classification criterion, before SVM modeling. The proposed data processing strategy allowed the classification of the wine sample set with accuracies up to 100% in both cross-validation and on an independent test set and highlighted the efficiency of 1H NMR and Raman data fusion as opposed to the use of a single-source data for differentiating wine concerning the cultivar and vintage.

The development of honey recognition models with broad applicability based on the association of isotope and elemental content with ANNs.

Honey adulteration represents a worldwide problem, driven by the illicit economic gain that producers, traders, or merchants pursue. Among the falsification methods that can unfairly influence the price is the incorrect declaration of the botanical origin and harvesting year. Therefore, the present study aimed to test the potential given by the application of Artificial Neural Networks (ANNs) for developing prediction models able to assess honey botanical origin and harvesting year based on isotope and elemental fingerprints. For each classification criterion, significant focus was dedicated to the data preprocessing phase to enhance the models' prediction capability. The obtained classification performances (accuracy scores >86% during the test phase) have highlighted the efficiency of ANNs for honey authentication as well as the feasibility of applying the developed classifiers for a large-scale application, supported by their ability to recognize the correct origin despite considerable variability in botanical source, geographical origin, and harvesting period.

Geographical Origin Authentication-A Mandatory Step in the Efficient Involvement of Honey in Medical Treatment.

Nowadays, in people's perceptions, the return to roots in all aspects of life is an increasing temptation. This tendency has also been observed in the medical field, despite the availability of high-level medical services with many years of research, expertise, and trials. Equilibrium is found in the combination of the two tendencies through the inclusion of the scientific experience with the advantages and benefits provided by nature. It is well accepted that the nutritional and medicinal properties of honey are closely related to the botanical origin of the plants at the base of honey production. Despite this, people perceive honey as a natural and subsequently a simple product from a chemical point of view. In reality, honey is a very complex matrix containing more than 200 compounds having a high degree of compositional variability as function of its origin. Therefore, when discussing the nutritional and medicinal properties of honey, the importance of the geographical origin and its link to the honey's composition, due to potential emerging contaminants such as Rare Earth Elements (REEs), should also be considered. This work offers a critical view on the use of honey as a natural superfood, in a direct relationship with its botanical and geographical origin.

Geographical and varietal origin differentiation of alcoholic beverages through the association between FT-Raman spectroscopy and advanced data processing strategies.

The present work aimed to test the efficiency of FT-Raman spectroscopy for fruit spirits discrimination by developing differentiation models based on two approaches, namely a supervised statistical method (Partial Least Squares Discriminant Analysis), and a Machine Learning technique (Support Vector Machines). For this purpose, a data set comprising 86 Romanian distillate samples was used, which aimed to be differentiated in terms of the raw material used for production (plum, apple, pear and grape) and county of origin (Cluj, Satu Mare and Salaj). Eight distinct preprocessing methods (autoscale, mean center, variance scaling, smoothing, 1st derivative, 2nd derivative, standard normal variate and Pareto) followed by a feature selection step were applied to identify the meaningful input data based on which the most efficient classification models can be constructed. Both types of models led to accuracy scores greater than 90% in differentiating the distillate samples in terms of geographical and botanical origin.

Botanical honey recognition and quantitative mixture detection based on Raman spectroscopy and machine learning.

The development of new approaches for honey recognition, based on spectroscopic techniques, presents a huge market potential especially because of the fast development of portable equipment. As an emerging approach, the association between Raman spectroscopy and Artificial Intelligence (i.e. Machine Learning algorithms) for food and beverages recognition starts to prove its efficiency, becoming an important candidate for the development of a practical application. Through this study, new recognition models for the rapid and efficient botanical differentiation of investigated honey varieties were developed, allowing the correct prediction of each type in a percentage better than 81%. The performances of the constructed models were expressed in terms of precision, sensitivity, and specificity. Moreover, through this approach, the detection of honey mixtures was possible to be made and an estimative percentage of the mixture components was obtained. Thus, the applicative potential of this new approach for honey recognition as well as a qualitative and quantitative estimation of the honey mixture was demonstrated.

Isotopic and Elemental Fingerprint of Edible Egg Parts-The Health Risk Assessment Based on Potentially Toxic Elements Content.

The present study investigated the isotopic and elemental profile (by IRMS and ICP-MS) of edible egg parts (29 egg whites and 29 yolks) mainly collected from Romania. In order to differentiate the egg white and yolk coming from different hen rearing systems (backyard and barn), Partial Least Square-Discriminant Analysis (PLS-DA) models were developed. The models' accuracies for the discrimination according to the hen growing system were 96% for egg white and 100% for egg yolk samples, respectively. Elements that proved to have the highest discrimination power for both egg white and yolk were the following: δ13C, Li, B, Mg, K, Ca, Mn, Fe, Co, Zn, Rb, Sr, Mo, Ba, La, Ce, and Pb. Nevertheless, the important compositional differentiation, in terms of essential mineral content, between the edible egg parts (egg white and egg yolk) were also pointed out. The estimated daily intake (EDI), the target hazard quotient (THQ) for Cr, Mn, Fe, Co, Ni, Cu, Zn, Se, Cd, Pb, and As, as well as the hazard index (HI) were used to assess non-carcinogenic human health risks from egg consumption. The obtained results showed no noticeable health risks related to egg consumption for humans from the point of view of the potentially toxic metals.

Differentiation of Transylvanian fruit distillates using supervised statistical tools based on isotopic and elemental fingerprint.

BACKGROUND: The spirit drinks industry is one of the largest in the world. Fruit distillates require adequate analysis methods combined with statistical tools to build differentiation models, according to distinct criteria (geographical and botanical origin, producer's fingerprint, respectively). Over time a database of alcoholic beverage fingerprints can be generated, being very important for product safety and authenticity control. RESULTS: To control the distillates' geographical origin, linear discriminant analysis (LDA) revealed that the cross-validation classification was correct for 88.2% of samples, but partial least squares discriminant analysis (PLS-DA) was slightly better suited for this purpose, with a correct classification rate of 91.2%. LDA effectiveness was proven for the trademark fingerprint differentiation, which was achieved at 93.5%, compared to 89.1% for PLS-DA. The principal predictors obtained by LDA were the same both for geographical origin and producer differentiation: B, δ13 C, Na, Cu, Ca and Be; highlighting the fact that in the production process of distillates each producer used fruits coming from the respective specific region. Through PLS-DA, some of the discrimination markers were the same for geographical origin and producer's identification, but others were completely specific: the rare earth elements Eu and Er only for geographical origin differentiation, and Cu solely as predictor for producer's identification. Regarding distillates' fruit variety, the correct discrimination rates of plum spirits from the rest were 84.2% for PLS-DA and 63% for LDA. CONCLUSION: LDA and PLS-DA were suitable for differentiation models development of fruits spirits according to geographical region, producer and fruit variety based on isotopic and elemental fingerprint. © 2022 Society of Chemical Industry.

Application of fuzzy algorithms in conjunction with 1 H-NMR spectroscopy to differentiate alcoholic beverages.

BACKGROUND: Recent statistics from the European Commission indicate that wine is one of the commodities most commonly subject to food fraud. In this context, the development of reliable classification models to differentiate alcoholic beverages requires, besides sensitive analytical tools, the use of the most suitable data-processing methods like those based on advanced statistical tools or artificial intelligence. RESULTS: The present study aims to establish a new, innovative approach for the differentiation of alcoholic beverages (wines and fruit distillates), which is able to increase the discrimination rate of the models that have been developed. A data dimensionality reduction step was applied to proton nuclear magnetic resonance (1 H-NMR) profiles. This stage consisted of the application of fuzzy principal component analysis (FPCA) prior to the development of classification models through discriminant analysis. The enhancement of the model's classification potential by the application of FPCA in comparison with principal component analysis (PCA) was discussed. CONCLUSION: The association of 1 H-NMR spectroscopy and an appropriate statistical approach provided a very effective tool for the differentiation of alcoholic beverages. To develop reliable metabolomic approaches for the differentiation of wines and fruit distillates, 1 H-NMR spectroscopic data were exploited in conjunction with fuzzy algorithms to reduce data dimensionality. The study proved the greater efficiency of using FPCA scores in comparison with those obtained through the widely applied PCA. The proposed approach enabled wines to be distinguished perfectly according to their geographical origins, cultivar, and vintage, and this could be used for wine classification. Moreover, 100% correctly classified samples were also achieved for the botanical and geographical differentiation of fruit distillates. © 2022 Society of Chemical Industry.

Machine Learning Approach to Comparing Fatty Acid Profiles of Common Food Products Sold on Romanian Market.

Food composition issues represent an increasing concern nowadays, in the context of diverse food commodity varieties. The contents and types of fatty acids are a constant preoccupation among consumers because of their reflections of nutrition and health problems. This study aims to find the best tool for the rapid and reliable identification of similarities and differences among several food items from a fatty acid profile perspective. An acknowledged GC-FID method was considered, while, for a better interpretation of the analytical results, machine learning algorithms were used. It was possible to develop a recognition model able to simultaneously differentiate, with an accuracy of 79.3%, nine product types using the bagged tree ensemble model. The low number of samples or some similarities among the classes could be responsible for the wrong assignments that occurred, especially in the biscuit, wafer and instant soup classes. Better accuracies values of 95, 86.1, and 97.8% were obtained when the products were grouped into three categories: (1) sunflower oil, mayonnaise, margarine, and cream cheese; (2) biscuits, cookies, margarine, and wafers; and (3) sunflower oil, chips, and instant soup.

Influence of Geographical Origin on Isotopic and Elemental Compositions of Pork Meat.

Pigs are a primary source of meat, accounting for over 30% of global consumption. Consumers' preferences are determined by health considerations, paying more attention to foodstuffs quality, animal welfare, place of origin, and swine feeding regime, and being willing to pay a higher price for a product from a certain geographical region. In this study, the isotopic fingerprints (δ2H, δ18O, and δ13C) and 29 elements of loin pork meat samples were corroborated with chemometric methods to obtain the most important variables that could classify the samples' geographical origin. δ2H and δ18O values ranged from -71.0 to -21.2‱, and from -9.3 to -2.8‱, respectively. The contents of macro- and micro-essential elements are presented in the following order: K > Na > Mg > Ca > Zn > Fe > Cu > Cr. The LDA model assigned in the initial classification showed 91.4% separation of samples, while for the cross-validation procedure, a percentage of 90% was obtained. δ2H, K, Rb, and Pd were identified as the most representative parameters to differentiate the pork meat samples coming from Romania vs. those from abroad. The mean values of metal concentrations were used to estimate the potential health risks associated with the consumption of pork meat The results showed that none of the analyzed metals (As, Cd, Sn, Pb, Cu, and Zn) pose a carcinogenic risk.

The Development of Honey Recognition Models Based on the Association between ATR-IR Spectroscopy and Advanced Statistical Tools.

The newly developed prediction models, having the aim to classify Romanian honey samples by associating ATR-FTIR spectral data and the statistical method, PLS-DA, led to reliable differentiations among the samples, in terms of botanical and geographical origin and harvesting year. Based on this approach, 105 out of 109 honey samples were correctly attributed, leading to true positive rates of 95% and 97% accuracy for the harvesting differentiation model. For the botanical origin classification, 83% of the investigated samples were correctly predicted, when four honey varieties were simultaneously discriminated. The geographical assessment was achieved in a percentage of 91% for the Transylvanian samples and 85% of those produced in other regions, with overall accuracy of 88% in the cross-validation procedure. The signals, based on which the best classification models were achieved, allowed the identification of the most significant compounds for each performed discrimination.

Isotopic and elemental characterization of Romanian pork meat in corroboration with advanced chemometric methods: A first exploratory study.

In this study 93 pork meat samples (tenderloin) were analyzed via isotope ratios mass spectrometry (δ2H, δ18O, δ13C) and inductively coupled plasma - Mass spectrometry (55 elements). The meat samples are coming from Romania and abroad. Those from Romania are originating from conventional farms and yard rearing system. The analytical results in conjunction with linear discriminant analysis (LDA) and artificial neural networks (ANNs) were used to assess: The geographical origin, and animal diet. The most powerful markers which could differentiate pork meat samples concerning the geographical origin were δ18O, terbium, and tin. The results of chemometric models showed that, along with 13C signature, rubidium concentration, and a few rare earth-elements (lanthanum, and cerium) were efficient to discriminate animal diet in a percent of 97.8% (initial classification) and 94.6% (cross-validation), respectively. Some of predictors for feeding regime differentiation by using LDA were identified also to be the best markers to distinguish corn-based diet by using ANNs (δ13C, Rb, La).

Authentication of Transylvanian Spirits Based on Isotope and Elemental Signatures in Conjunction with Statistical Methods.

The potential association between stable isotope ratios of light elements and mineral content, in conjunction with unsupervised and supervised statistical methods, for differentiation of spirits, with respect to some previously defined criteria, is reviewed in this work. Thus, based on linear discriminant analysis (LDA), it was possible to differentiate the geographical origin of distillates in a percentage of 96.2% for the initial validation, and the cross-validation step of the method returned 84.6% of correctly classified samples. An excellent separation was also obtained for the differentiation of spirits producers, 100% in initial classification, and 95.7% in cross-validation, respectively. For the varietal recognition, the best differentiation was achieved for apricot and pear distillates, a 100% discrimination being obtained in both classifications (initial and cross-validation). Good classification percentages were also obtained for plum and apple distillates, where models with 88.2% and 82.4% in initial and cross-validation, respectively, were achieved for plum differentiation. A similar value in the cross-validation procedure was reached for the apple spirits. The lowest classification percent was obtained for quince distillates (76.5% in initial classification followed by 70.4% in cross-validation). Our results have high practical importance, especially for trademark recognition, taking into account that fruit distillates are high-value commodities; therefore, the temptation of "fraud", i.e., by passing regular distillates as branded ones, could occur.

Applications of emerging stable isotopes and elemental markers for geographical and varietal recognition of Romanian and French honeys.

The research towards the identification of new authenticity markers is crucial to fight against fraudulent activities on honey, one of the top ten most falsified food commodities. This work proposes an association of stable isotopes and elemental content as markers for honey authentication, with respect to its floral and geographical origin. Emerging markers like isotopic signature of honey water alongside with carbon and hydrogen isotopic ratios of ethanol obtained from honey fermentation and Rare Earth Elements, were used to develop new recognition models. Thus, the efficiency of the discrimination potential of these emerging markers was discussed individually and in association. This approach proved its effectiveness for geographical differentiation (>98%) and the role of the emerging markers in these classifications was an essential one, especially of: (D/H)I, δ2H, δ18O, La, Ce and Pr. Floral recognition was realized in a lower percentage revealing the suitability of these markers mainly for geographical classification.

Application of Raman spectroscopy and Machine Learning algorithms for fruit distillates discrimination.

Through this pilot study, the association between Raman spectroscopy and Machine Learning algorithms were used for the first time with the purpose of distillates differentiation with respect to trademark, geographical and botanical origin. Two spectral Raman ranges (region I-200-600 cm-1 and region II-1200-1400 cm-1) appeared to have the higher discrimination potential for the investigated distillates. The proposed approach proved to be a very effective one for trademark fingerprint differentiation, a model accuracy of 95.5% being obtained (only one sample was misclassified). A comparable model accuracy (90.9%) was achieved for the geographical discrimination of the fruit spirits which can be considered as a very good one taking into account that this classification was made inside Transylvania region, among neighbouring areas. Because the trademark fingerprint is the prevailing one, the successfully distillate type differentiation, with respect to the fruit variety, was possible to be made only inside of each producing entity.

Fuzzy Divisive Hierarchical Associative-Clustering Applied to Different Varieties of White Wines According to Their Multi-Elemental Profiles.

Wine data are usually characterized by high variability, in terms of compounds and concentration ranges. Chemometric methods can be efficiently used to extract and exploit the meaningful information contained in such data. Therefore, the fuzzy divisive hierarchical associative-clustering (FDHAC) method was efficiently applied in this study, for the classification of several varieties of Romanian white wines, using the elemental profile (concentrations of 30 elements analyzed by ICP-MS). The investigated wines were produced in four different geographical areas of Romania (Transylvania, Moldova, Muntenia and Oltenia). The FDHAC algorithm provided not only a fuzzy partition of the investigated white wines, but also a fuzzy partition of considered characteristics. Furthermore, this method is unique because it allows a 3D bi-plot representation of membership degrees corresponding to wine samples and elements. In this way, it was possible to identify the most specific elements (in terms of highest, smallest or intermediate concentration values) to each fuzzy partition (group) of wine samples. The chemical elements that appeared to be more powerful for the differentiation of the wines produced in different Romanian areas were: K, Rb, P, Ca, B, Na.

Raman spectroscopy and machine-learning for edible oils evaluation.

Because of the important advantages as rapidity, cost effectiveness and no sample preparation necessity, encountered in most of the cases, Raman spectroscopy gained more and more attention during the last years with regard to its application in food and beverages authenticity. Vegetable cold-pressed oils obtained from: sesame, hemp, walnut, linseed, pumpkin and sea buckthorn have gained increased attention in consumer interest due to their high nutrient value and health benefits. The high commercial value of these, brought the temptation from some unfair producers and sellers to gain an illegal profit by replacing the raw material of these oils by cheaper ones (i.e. sunflower). Here a new approach based on the rapid processing of Raman spectra using Machine Learning algorithms, for edible oil authentication was developed and successfully tested. Through this approach, not only the adulteration detection was achieved but also an initial estimation of its magnitude.

Differentiation of tomatoes based on isotopic, elemental and organic markers.

In this study, 41 tomato samples were investigated by means of stable isotope ratios (δ13C, δ18O and δ2H), elemental content, phenolic compounds and pesticides in order to classify them, according to growing conditions and geographical origin. Using investigated parameters, stepwise linear discriminant analysis was applied and the differences that occurred between tomato samples grown in greenhouses compared to those grown on field, and also between Romanian and abroad purchased samples were pointed out. It was shown that Ti, Ga, Te, δ2H and δ13C content were able to differentiate Romanian tomato samples from foreign samples, whereas Al, Sc, Se, Dy, Pb, δ18O, 4,4'-DDT could be used as markers for growing regime (open field vs. greenhouse). For the discrimination of different tomato varieties (six cherry samples and fourteen common sorts) grown in greenhouse, phenolic compounds of 20 samples were determined. In this regard, dihydroquercetin, caffeic acid, chlorogenic acid, rutin, rosmarinic acid, quercetin and naringin were the major phenolic compounds detected in our samples. The phenolic profile showed significant differences between cherry tomato and common tomato. The contents of the chlorogenic acid and rutin were significantly higher in the cherry samples (90.27-243.00 µg/g DW and 160.60-433.99 µg/g DW respectively) as compared to common tomatoes (21.30-88.72 µg/g DW and 24.84-110.99 µg/g DW respectively). The identification of dihydroquercetin is of particular interest, as it had not been reported previously in tomato fruit.

Isotopic and elemental profiling alongside with chemometric methods for vegetable differentiation.

In this study, three chemometric models for vegetables growing system (field versus greenhouse), geographical origin and species attribution using stable isotope (δ13C, δ18O, δ2H) and elemental fingerprints of 101 samples (54 squashes and 47 radishes) commercialized on Romanian market were developed. These models were constructed and validated through linear discriminant analysis. Initial validations of 94.4% and 83% were obtained for squash and radish growing systems, respectively, such that one squash and four radish samples declared to be grown in the field were attributed to the greenhouse group. For this purpose, the most powerful differentiation markers appeared to be Sn and δ13C for radishes, and Sn, Cu for squashes. Regarding the vegetable origin, four samples, initially considered to originate from Romania (95% for initial classification) were attributed to the foreign group in the cross-validation procedure (93.1%). Romanian radishes and squashes were characterized by a higher content of Na and Cu, respectively, compared with foreign samples, while the mean values for Zn, Sr, Zr and Co concentrations were found to be higher for the vegetables from abroad.

An optimized green preparation method for the successful application of Raman spectroscopy in honey studies.

Raman spectroscopy represents an emerging technique for food authentication being a fast, reliable analytical method, requiring a minimum sample preparation step. Anyway, as in the case of any analytical techniques, there are some limitations which need to be properly assessed before applying this method in honey authentication. In this regard, the aim of this study consisted in the development of a simple working protocol, for honey sample preparation, which can simultaneously overcome the main limitations of Raman spectroscopy in honey studies, such as crystallization and fluorescence. Thus, in this work, a new green sample preparation method is proposed, discussed and its robustness is tested. It has been demonstrated that through honey dilution, in distilled water, reliable and reproductible spectra could be obtained, allowing the investigation of different types of honey. The main advantage of the method consists in the simultaneously overcoming of the most significant limitations of Raman spectroscopy employment in honey studies, such as crystallization and fluorescence, by a simple 1:1 w/v dilution of honey in distilled water.