Application of random forests to select premium quality vegetable oils by their fatty acid composition

Detection of camellia oil adulteration based on near-infrared spectroscopy and smartphone combined with deep learning and multimodal fusion

Camellia oil (CO) is known for its nutritional value and health benefits, but its high price makes it susceptible to adulteration. This study developed a binary adulteration system for CO in response to the adulteration of rapeseed oil (RO) into CO that been observed in the market. A total of 243 oil samples adulterated with various concentrations of RO were prepared. The spectral information of the adulterated oil samples was obtained using near-infrared (NIR) spectroscopy. Additionally, visual data obtained from smartphone-captured images and videos were analysed. Deep-learning models trained on video data reached the highest accuracy of 96.30 %. To improve detection accuracy, a multimodal approach was adopted by combing spectral and visual data. Generally, this study presented a novel method for detecting the authenticity of CO in real time, providing technical support to address increasingly serious food safety concerns and laying the foundation for future rapid online detection using smartphones.

In silico identification for flavor antioxidant compounds in Chrysanthemi flos uncovers the interactions between saccharides and secondary metabolites

Secondary metabolites and saccharides are responsible for antioxidant activity and flavor of Chrysanthemi flos (CF). However, the flavor antioxidant compounds of CF and their intermolecular interactions remain unclear. Here, we primarily employed in silico methods to identify CF antioxidants. After characterizing by physicochemical properties, FT-NIR and HPLC fingerprint, the "spectrum-effect" fusion correlation was established to select the spectral features of CF antioxidants. Quercetagitrin (QU), chlorogenic acid (CA) and saccharides fragments were clarified based on their characteristic spectrum. The antioxidant efficacy as well as the sweet and bitter taste of these compounds were verified by molecular docking. Quantum chemical calculations demonstrated that non-covalent interactions dominant facilitated the stable existence of CF antioxidants. The most significant binding types between CA, QU and saccharides fragments were hydrogen bonding. These results indicate a novel approach and theoretical support to discovery of new information pertinent to the bioactive compounds related to CF or other tea.

Geographical Origin Traceability of Navel Oranges Based on Near-Infrared Spectroscopy Combined with Deep Learning

The quality and price of navel oranges vary depending on their geographical origin, thus providing a financial incentive for origin fraud. To prevent this phenomenon, it is necessary to explore a fast, non-destructive, and precise method for tracing the origin of navel oranges. In this study, a total of 490 Newhall navel oranges were selected from five major production regions in China, and the diffuse reflectance near-infrared spectrum in 4000-10,000 cm-1 were non-invasively collected. We examined seven preprocessing techniques for the spectra, including Savitzky-Golay (SG) smoothing, first derivative (FD), multiplicative scattering correction (MSC), combinations of SG with MSC (SG+MSC), SG with FD (SG+FD), MSC with FD (MSC+FD), and three combined (SG+MSC+FD). A one-dimensional convolutional neural network (1DCNN) deep learning model for geographical origin tracing of navel orange was established, and five machine learning algorithms, i.e., partial least squares discriminant analysis (PLS-DA), linear discriminant analysis (LDA), support vector machine (SVM), random forest (RF), and back-propagation neural network (BPNN), were compared with 1DCNN. The results show that the 1DCNN model based on the SG+FD preprocessing method achieved the optimal performance for the testing set, with prediction accuracy, precision, recall, and F1-score of 97.92%, 98%, 97.95%, and 97.90%, respectively. Therefore, NIRS combined with deep learning has a significant research and application value in the rapid, nondestructive, and accurate geographical origin traceability of agricultural products.

Integrating near-infrared hyperspectral imaging with machine learning and feature selection: Detecting adulteration of extra-virgin olive oil with lower-grade olive oils and hazelnut oil

Detecting adulteration in extra virgin olive oil (EVOO) is particularly challenging with oils of similar chemical composition. This study applies near-infrared hyperspectral imaging (NIR-HSI) and machine learning (ML) to detect EVOO adulteration with hazelnut, refined olive, and olive pomace oils at various concentrations (1%, 5%, 10%, 20%, 40%, and 100% m/m). Savitzky-Golay filtering, first and second derivatives, multiplicative scatter correction (MSC), standard normal variate (SNV), and their combinations were used to preprocess the spectral data, with Principal Component Analysis (PCA) reducing dimensionality. Classification was performed using Partial Least Squares-Discriminant Analysis (PLS-DA) and ML algorithms, including k-Nearest Neighbors (k-NN), Naïve Bayes, Random Forest (RF), Support Vector Machine (SVM), and Artificial Neural Networks (ANN). PLS-DA, k-NN, RF, SVM, NB, and ANN models achieved accuracy rates of 97.0-99.0%, 96.2-100%, 96.5-100%, 98.6-99.5%, 93.9-99.7%, and 99.2-100%, respectively, in discriminating between pure EVOO, adulterants, and adulterated oils. PLS-DA, RF, SVM, and ANN significantly outperformed Naïve Bayes (p < 0.05) in binary classification, with Matthews correlation coefficient (MCC) values exceeding 0.90. All the binary classifiers except Naïve Bayes, when coupled with SNV/MSC, Savitzky-Golay smoothing and derivatives, consistently achieved perfect scores (1.0) for accuracy, sensitivity, specificity, F1 score, precision, and MCC in distinguishing pure EVOO from adulterated oils. No significant differences (p > 0.05) in model performance were found between those using full spectra and those based on key variable selection. However, PLS-DA and ANN significantly outperformed k-NN, RF, and SVM (p < 0.05), with MCC values ranging from 0.95 to 1.00, indicating superior classification performance. These findings demonstrate that combining NIR-HSI with machine learning, along with key variable selection, potentially offers an effective, non-destructive solution for detecting adulteration in EVOO and combating fraud in the olive oil industry.

Unraveling the Metabolomics Mysteries in Camellia Oil: From Cognition to Application

Camellia oil is a high-value edible seed oil, recommended by the Food and Agriculture Organization (FAO). It is essential to develop accurate and rapid analytical methods to authenticate camellia oil due to its susceptibility to adulteration. Recently, hyphenated chromatography-mass spectrometry, especially high-resolution mass spectrometry using chemometrics, has become a promising platform for the identification of camellia oil. Based on the compositional analysis, the fatty acid, sterol, phenol, and tocopherol profiles (or fingerprints) were utilized as predictor variables for assessing authenticity. The review systematically summarizes the workflow of chromatography-mass spectrometry technologies and comprehensively investigates recent metabolomic applications combined with chemometrics for camellia oil authentication. Metabolomics has significantly improved our understanding of camellia oil composition at the molecular level, contributing to its identification and full characterization. Hence, its integration with standard analytical methods is essential to enhance the tools available for public and private laboratories to assess camellia oil authenticity. Integrating metabolomics with artificial intelligence is expected to accelerate drug discovery by identifying new metabolic pathways and biomarkers, promising to revolutionize medicine.

Raman spectroscopy and multivariate analysis for the waste and edible vegetable oil classification

Twelve samples of waste cooking oil (WCO) were prepared by four different deep-frying procedures. The edible and the waste oil samples were characterised by Raman spectroscopy, revealing few and almost negligible differences between them. Therefore, the possibility of classifying the different groups of samples by extracting valuable data from the Raman spectra through statistical multivariate analysis was explored. Even if the number of samples was not enough to draw definitive conclusions, unsupervised principal component analysis (PCA) and supervised partial least square discriminant analysis (PLS-DA) conducted on the raw Raman signals, allowed to distinguish within edible and waste vegetable oil, and to select the most relevant combination of variables associated with each family. Using sparse partial least square discriminant analysis (S-PLS-DA), we determined a chemical fingerprint characteristic of each sample by creating a Variable In Projection (VIP) plot. The methodology herein presented could find relevant application in the detection of waste adulteration in vegetable oils sold for industrial purposes other than food.

The Ability of Different Tea Tree Germplasm Resources in South China to Aggregate Rhizosphere Soil Characteristic Fungi Affects Tea Quality

It is generally recognized that the quality differences in plant germplasm resources are genetically determined, and that only a good "pedigree" can have good quality. Ecological memory of plants and rhizosphere soil fungi provides a new perspective to understand this phenomenon. Here, we selected 45 tea tree germplasm resources and analyzed the rhizosphere soil fungi, nutrient content and tea quality. We found that the ecological memory of tea trees for soil fungi led to the recruitment and aggregation of dominant fungal populations that were similar across tea tree varieties, differing only in the number of fungi. We performed continuous simulation and validation to identify four characteristic fungal genera that determined the quality differences. Further analysis showed that the greater the recruitment and aggregation of Saitozyma and Archaeorhizomyces by tea trees, the greater the rejection of Chaetomium and Trechispora, the higher the available nutrient content in the soil and the better the tea quality. In summary, our study presents a new perspective, showing that ecological memory between tea trees and rhizosphere soil fungi leads to differences in plants' ability to recruit and aggregate characteristic fungi, which is one of the most important determinants of tea quality. The artificial inoculation of rhizosphere fungi may reconstruct the ecological memory of tea trees and substantially improve their quality.

Recruitment and Aggregation Capacity of Tea Trees to Rhizosphere Soil Characteristic Bacteria Affects the Quality of Tea Leaves

There are obvious differences in quality between different varieties of the same plant, and it is not clear whether they can be effectively distinguished from each other from a bacterial point of view. In this study, 44 tea tree varieties (Camellia sinensis) were used to analyze the rhizosphere soil bacterial community using high-throughput sequencing technology, and five types of machine deep learning were used for modeling to obtain characteristic microorganisms that can effectively differentiate different varieties, and validation was performed. The relationship between characteristic microorganisms, soil nutrient transformation, and tea quality formation was further analyzed. It was found that 44 tea tree varieties were classified into two groups (group A and group B) and the characteristic bacteria that distinguished them came from 23 genera. Secondly, the content of rhizosphere soil available nutrients (available nitrogen, available phosphorus, and available potassium) and tea quality indexes (tea polyphenols, theanine, and caffeine) was significantly higher in group A than in group B. The classification result based on both was consistent with the above bacteria. This study provides a new insight and research methodology into the main reasons for the formation of quality differences among different varieties of the same plant.

Machine Learning Application in Horticulture and Prospects for Predicting Fresh Produce Losses and Waste: A Review

The current review examines the state of knowledge and research on machine learning (ML) applications in horticultural production and the potential for predicting fresh produce losses and waste. Recently, ML has been increasingly applied in horticulture for efficient and accurate operations. Given the health benefits of fresh produce and the need for food and nutrition security, efficient horticultural production and postharvest management are important. This review aims to assess the application of ML in preharvest and postharvest horticulture and the potential of ML in reducing postharvest losses and waste by predicting their magnitude, which is crucial for management practices and policymaking in loss and waste reduction. The review starts by assessing the application of ML in preharvest horticulture. It then presents the application of ML in postharvest handling and processing, and lastly, the prospects for its application in postharvest loss and waste quantification. The findings revealed that several ML algorithms perform satisfactorily in classification and prediction tasks. Based on that, there is a need to further investigate the suitability of more models or a combination of models with a higher potential for classification and prediction. Overall, the review suggested possible future directions for research related to the application of ML in postharvest losses and waste quantification.

Evaluation of different regression models for detection of adulteration of mustard and canola oil with argemone oil using fluorescence spectroscopy coupled with chemometrics

Mustard and canola oils are commonly used cooking oils in Asian countries such as India, Nepal, and Bangladesh, making them prone to adulteration. Argemone is a well-known adulterant of mustard oil, and its alkaloid sanguinarine has been linked with health conditions such as glaucoma and dropsy. Utilising a non-destructive spectroscopic method coupled with a chemometric approach can serve better for the detection of adulterants. This work aimed to evaluate the performance of various regression algorithms for the detection of argemone in mustard and canola oils. The spectral dataset was acquired from fluorescence spectrometer analysis of pure as well as adulterated mustard and canola oils with some local and commercial samples also. The prediction performance of the eight regression algorithms for the detection of adulterants was evaluated. Extreme gradient boosting regressor (XGBR), Category gradient boosting regressor (CBR), and Random Forest (RF) demonstrate potential for predicting adulteration levels in both oils with high R2 values.

Detection of goat herding impact on vegetation cover change using multi-season, multi-herd tracking and satellite imagery

The frequency and severity of Mediterranean forest fires are expected to worsen as climate change progresses, heightening the need to evaluate understory fuel management strategies as rigorously as possible. Prescribed small-ruminant foraging is considered a sustainable, cost-effective strategy, but demonstrating a link between animal presence and vegetation change is challenging. This study tested whether the effect of small-ruminant herd presence in Mediterranean woodlands can be detected by integrating remote sensing and herd tracking at the landscape scale. The daily foraging routes of seven shepherded goat herds that exploited a 100-km2 forested area of the Judean Hills, Israel, were tracked over six years using GPS (Global Positioning System) collars. Herd locations were converted to stocking rates, with units of animal-presence-days per unit area per defined time period, and mapped at a spatial resolution of 10 m. We estimated pixel-level vegetation cover change based on a time series of 63 monthly Landsat-8 images expressed as the normalized soil-adjusted vegetation index (SAVI). Spatiotemporal trend analysis assessed the magnitude and direction of change, and a random forest machine-learning algorithm estimated the relative impact on vegetation cover change of environmental factors as well as the herd-related factors of stocking rate that accrued over six years and distance to the closest corral. The last two factors were among the most influential factors determining vegetation cover change in the regional and individual-herd analyses. In some respects, the permanent herds differed in their spatial pattern of stocking rate from the mobile herds that periodically relocated their night corral throughout the year, but stocking rate scaled logarithmically for all herds individually and combined. The combination of multi-season GPS tracking, remote sensing, and machine-learning techniques, applied at a regional scale, detected herd impacts on vegetation cover trends, consistent with livestock foraging being an effective tool for fuel reduction in Mediterranean woodlands.

Discrimination of Port wines by style and age using chromatic characteristics, phenolic, and pigment composition

The colour of the different Port wine styles and indication of age (IOA) categories is a distinctive quality parameter influenced by the grapes and ageing process. The impact of Port wine styles and IOA on phenolic composition is mostly unknown. This work aims to study the chromatic characteristics (CIELab) and their relation with the phenolic composition of White, Tawny, and Ruby Port wines and evaluate the feasibility of its utilisation for their discrimination. Port wine styles and IOA categories can be discriminated by their chromatic characteristics, using different data analysis models. The higher b* values, corresponding to the brownish/yellowish colour of Tawny and White Ports belonging to higher IOA categories, seem more related to the sugar browning than the oxidative change in phenolic compounds. However, this last process is essential for the red colour (a*) decrease of Tawny Port wines with higher IOA.

A Study on Origin Traceability of White Tea (White Peony) Based on Near-Infrared Spectroscopy and Machine Learning Algorithms

Identifying the geographical origins of white tea is of significance because the quality and price of white tea from different production areas vary largely from different growing environment and climatic conditions. In this study, we used near-infrared spectroscopy (NIRS) with white tea (n = 579) to produce models to discriminate these origins under different conditions. Continuous wavelet transform (CWT), min-max normalization (Minmax), multiplicative scattering correction (MSC) and standard normal variables (SNV) were used to preprocess the original spectra (OS). The approaches of principal component analysis (PCA), linear discriminant analysis (LDA) and successive projection algorithm (SPA) were used for features extraction. Subsequently, identification models of white tea from different provinces of China (DPC), different districts of Fujian Province (DDFP) and authenticity of Fuding white tea (AFWT) were established by K-nearest neighbors (KNN), random forest (RF) and support vector machine (SVM) algorithms. Among the established models, DPC-CWT-LDA-KNN, DDFP-OS-LDA-KNN and AFWT-OS-LDA-KNN have the best performances, with recognition accuracies of 88.97%, 93.88% and 97.96%, respectively; the area under curve (AUC) values were 0.85, 0.93 and 0.98, respectively. The research revealed that NIRS with machine learning algorithms can be an effective tool for the geographical origin traceability of white tea.

The phytochemical components of walnuts and their application for geographical origin based on chemical markers

Place of origin has an important influence on walnut quality and commercial value, which results in the requirement of rapid geographical traceability method. Thus, a method for geographical origin identification of walnuts on the basis of nutritional quality of walnut from China was conducted. The concentrations of 43 phytochemical components were analyzed in walnut samples from five different walnut-producing regions of China. Based on 14 chemical markers selected by the Random Forest method from these phytochemical components, a new discriminant model for geographical origin was built, with the corresponding correct classification rate of 99.3%. In addition, the quantitative quality differences of walnuts from five regions were analyzed, with the values of 0.17-1.43. Moreover, the top three chemical markers for the geographical origin discriminant analysis were Mo, V and stearic acid, with the contribution rates of 26.8%, 18.9% and 10.9%. This study provides a potentially viable method for application in the food authentication.

Variations in Fatty Acids Affected Their Derivative Volatiles during Tieguanyin Tea Processing

Fatty acids (FAs) are important precursors of oolong tea volatile substances, and their famous derivatives have been shown to be the key aroma components. However, the relationship between fatty acids and their derivatives during oolong tea production remains unclear. In this study, fresh Tieguanyin leaves were manufactured into oolong tea and green tea (control), and fatty acids and fatty acid-derived volatiles (FADV) were extracted from processed samples by the sulfuric acid–methanol method and solvent-assisted flavor evaporation (SAFE), respectively. The results showed that unsaturated fatty acids were more abundant than saturated fatty acids in fresh leaves and decreased significantly during tea making. Relative to that in green tea, fatty acids showed larger variations in oolong tea, especially at the green-making stage. Unlike fatty acids, the FADV content first increased and then decreased. During oolong tea manufacture, FADV contents were significantly and negatively correlated with total fatty acids; during the green-making stage, methyl jasmonate (MeJA) content was significantly and negatively correlated with abundant fatty acids except steric acid. Our data suggest that the aroma quality of oolong tea can be improved by manipulating fatty acid transformation.

Maillard reaction products and guaiacol as production process and raw material markers for the authentication of sesame oil

BACKGROUND

Sesame oil has an excellent flavor and is widely appreciated. It has a higher price than other vegetable oils because of the high price of its raw materials, and different processing techniques also result in products of different quality levels, which can command different prices. In the market, there is a persistent problem of adulteration of sesame oil, driven by economic interests. The screening of volatile markers used to distinguish the authenticity of sesame oil raw materials and production processes is therefore very important.

RESULTS

In this work, six markers related to the production processes and raw materials of sesame oil were screened by gas chromatography-tandem mass spectrometry (GC-MS/MS) combined with chemometric analysis. They were 3-methyl-2-butanone, 2-ethyl-5-methyl-pyrazine, guaiacol, 2,6-dimethyl-pyrazine, 5-methyl furfural, and ethyl-pyrazine. The concentration of these markers in sesame oil is between 10 and1000 times that found in other vegetable oils. However, only 3-methyl-2-butanone and 2-ethyl-5-methyl-pyrazine differed significantly as the result of the use of different production processes. Except for guaiacol, which was mainly derived from raw materials, the other five compounds mentioned above all result from the Maillard reaction during thermal processing. The six compounds mentioned above are sufficient to distinguish fraud involving sesame oil raw materials and production processes, and can identify accurately adulteration levels of 30% concentration.

CONCLUSION

In this study, the classification markers can identify the adulteration of sesame oil accurately. These six compounds are therefore important for the authenticity of sesame oil and provide a theoretical basis for the rapid and accurate identification of the authenticity of sesame oil. © 2021 Society of Chemical Industry.

Flood Risk Mapping by Remote Sensing Data and Random Forest Technique

Detecting effective parameters in flood occurrence is one of the most important issues that has drawn more attention in recent years. Remote Sensing (RS) and Geographical Information System (GIS) are two efficient ways to spatially predict Flood Risk Mapping (FRM). In this study, a web-based platform called the Google Earth Engine (GEE) (Google Company, Mountain View, CA, USA) was used to obtain flood risk indices for the Galikesh River basin, Northern Iran. With the aid of Landsat 8 satellite imagery and the Shuttle Radar Topography Mission (SRTM) Digital Elevation Model (DEM), 11 risk indices (Elevation (El), Slope (Sl), Slope Aspect (SA), Land Use (LU), Normalized Difference Vegetation Index (NDVI), Normalized Difference Water Index (NDWI), Topographic Wetness Index (TWI), River Distance (RD), Waterway and River Density (WRD), Soil Texture (ST]), and Maximum One-Day Precipitation (M1DP)) were provided. In the next step, all of these indices were imported into ArcMap 10.8 (Esri, West Redlands, CA, USA) software for index normalization and to better visualize the graphical output. Afterward, an intelligent learning machine (Random Forest (RF)), which is a robust data mining technique, was used to compute the importance degree of each index and to obtain the flood hazard map. According to the results, the indices of WRD, RD, M1DP, and El accounted for about 68.27 percent of the total flood risk. Among these indices, the WRD index containing about 23.8 percent of the total risk has the greatest impact on floods. According to FRM mapping, about 21 and 18 percent of the total areas stood at the higher and highest risk areas, respectively.

The use of analytical techniques coupled with chemometrics for tracing the geographical origin of oils: A systematic review (2013-2020)

The global market for imported, high-quality priced foods has grown dramatically in the last decade, as consumers become more conscious of food originating from around the world. Many countries require the origin label of food to protect consumers need about true characteristics and origin. Regulatory authorities are looking for an extended and updated list of the analytical techniques for verification of authentic oils and to support law implementation. This review aims to introduce the efforts made using various analytical tools in combination with the multivariate analysis for the verification of the geographical origin of oils. The popular analytical tools have been discussed, and scientometric assessment that underlines research trends in geographical authentication and preferred journals used for dissemination has been indicated. Overall, we believe this article will be a good guideline for food industries and food quality control authority to assist in the selection of appropriate methods to authenticate oils.

Characterization of iron walnut in different regions of China based on phytochemical composition

Little is known about the phytochemical composition of iron walnuts. Differences in the geographical origin of iron walnuts associated with economic benefits should also be examined. In this study, the phytochemical composition (fatty acids, Vitamin E, total polyphenols and flavonoids, amino acids, and minerals) of iron walnuts in China was investigated. The results showed that there were significant differences (p < 0.05) in the phytochemical composition of iron walnut oils and flours from different regions. Positive (r > 0.5, p < 0.05) and negative (r < - 0.5, p < 0.05) correlations were found between amino acids/minerals and amino acids/oleic acid, with the highest correlation coefficient (r = 0.742, p < 0.05) between Cu and tyrosine. In addition, based on the 12 phytochemical fingerprints selected by random forest, a geographical-origin identification model for iron walnuts was established, with a corresponding correct classification rate of 96.6%. The top three phytochemical fingerprints for the geographical-origin identification of iron walnut were microelements, macroelements, and antioxidant composition, with contribution rates of 61.7%, 18.1%, and 9.9%, respectively.

Detection of camellia oil adulteration using chemometrics based on fatty acids GC fingerprints and phytosterols GC-MS fingerprints

The fatty acid, squalene, and phytosterols, coupled to chemometrics were utilized to detect the adulteration of camellia oil (CAO) with palm superolein (PAO), refined olive oil (ROO), high oleic- sunflower oil (HO-SUO), sunflower oil (SUO), corn oil (COO), rice bran oil (RBO), rice oil (RIO), peanut oil (PEO), sesame oil (SEO), soybean oil (SOO), and rapeseed oil (RAO). CAO was characterized with higher triterpene alcohols, thus differentiated from other vegetable oils in principle component analysis (PCA). Using partial least squares-discriminant analysis (PLS-DA), CAO adulterated with PAO, ROO, HO-SUO, SUO, COO, RBO, RIO, PEO, SEO, SOO, RAO (5%-100%, w/w), could be classified, especially higher than 92.31% of the total discrimination accuracy, at an adulterated ratio above 30%. With less than 22 potential key markers selected by the variable importance in projection (VIP), the optimized PLS models were confirmed to be accurate for the adulterated level prediction in CAO.

Fatty Acid Profile of Lipid Fractions of Mangalitza (Sus scrofa domesticus) from Northern Romania: A GC-MS-PCA Approach

Mangalitza pig (Sus scrofa domesticus) becomes more popular in European countries. The goal of this study was to evaluate the fatty acid profile of the raw and thermally processed Mangalitza hard fat from Northern Romania. For the first time, the gas chromatography-mass spectrometry-Principal component analysis technique (GC-MS-PCA)—was applied to evaluate the dissimilarity of Mangalitza lipid fractions. Three specific layers of the hard fat of Mangalitza from Northern Romania were subjected to thermal treatment at 130 °C for 30 min. Derivatized samples were analyzed by GC-MS. The highest relative content was obtained for oleic acid (methyl ester) in all hard fat layers (36.1–42.4%), while palmitic acid was found at a half (21.3–24.1%). Vaccenic or elaidic acids (trans) were found at important concentrations of 0.3–4.1% and confirmed by Fourier-transform infrared spectroscopy. These concentrations are consistently higher in thermally processed top and middle lipid layers, even at double values. The GC-MS-PCA coupled technique allows us to classify the unprocessed and processed Mangalitza hard fat specific layers, especially through the relative concentrations of vaccenic/elaidic, palmitic, and stearic acids. Further studies are needed in order to evaluate the level of degradation of various animal fats by the GC-MS-PCA technique.

Comparative study and quality evaluation regarding morphology characters, volatile constituents, and triglycerides in seeds of five species used in traditional Chinese medicine

Volatile compounds (VCs) and triglycerides (TGs) are the primary groups of constituents in the fruits of five well-known species used in traditional Chinese medicine (TCM), viz. Alpinia oxyphylla Miq. (AO), Alpinia katsumadai Hayata (AK), Amomum villosum Lour. (FAL), Amomum villosum Lour. var. xanthioides T. L. Wu et Senjen (FALX), and Amomum longiligulare T. L. Wu (FALO). The fruits of these species are morphologically similar and commonly used in both foods and TCM. Each species is purportedly endowed with different medicinal properties. Efficient and environmentally friendly methods are desirable for the quality control of these species. The current study attempted to establish both comprehensive profiles and quality standards for the five TCM species. External morphology characters were provided to distinguish 18 fruit samples belonging to the five species, which were collected from different geographical regions of China. The VCs of each sample were analyzed by SPME GC/Q-ToF. The identification of marker compounds from each species allowed for the differentiation of the fruits from the five plants. Characterization and quantification of 21 TGs were achieved using SFC/MS with an analysis time of less than 15 min. The complex TGs were unambiguously identified using the MS detection with correct attribution of the acyl group to the sn-2 position. Moreover, the quantification of TGs was improved by using reference standards whenever possible or a single standard strategy to determine multiple TGs. The validity of the proposed SFC/MS method was assessed by analyzing fatty acids from the hydrolysis and transesterification products of the same sample set using GC/MS. The quantification results from both TGs and fatty acids were consistent, and were further substantiated by chemometric analysis. To our knowledge, this is the first comprehensive study utilizing the morphology, VCs, and TGs for quality evaluation purpose of these five TCM species.

Comparison of different classification algorithms to identify geographic origins of olive oils

Research on investigation and determination of geographic origins of olive oils is increased by consumers' demand to authenticated olive oils. Classification algorithms which are machine learning methods can be employed for the authentication of olive oils. In this study, different classification algorithms were evaluated to reveal the most accurate one for authentication of Turkish olive oils. BayesNet, Naive Bayes, Multilayer Perception, IBK, Kstar, SMO, Random Forest, J48, LWL, Logistic Regression, Simple Logistic, LogitBoost algorithms were implemented on 61 chemical analysis parameters of 49 olive oil samples from 6 different locations at Western Turkey. These 61 parameters were obtained from five different chemical analyses which are stable carbon isotope ratio, trace elements, sterol compositions, FAMEs and TAGs. This study is the most comprehensive study to determine the geographical origin of Turkish olive oils in terms of these mentioned features. Classification performances of the algorithms were compared using accuracy, specificity and sensitivity metrics. Random Forest, BayesNet, and LogitBoost algorithms were found as the best classification algorithms for authentication of Turkish olive oils. Using the classification model in this study, geographic origin of an unknown olive oil can be predicted with high accuracy. Besides, similar models can be developed to obtain useful information for authentication of other food products.

Characterization, Quantification and Quality Assessment of Avocado (Persea americana Mill.) Oils

Avocado oil is prized for its high nutritional value due to the substantial amounts of triglycerides (TGs) and unsaturated fatty acids (FAs) present. While avocado oil is traditionally extracted from mature fruit flesh, alternative sources such as avocado seed oil have recently increased in popularity. Unfortunately, sufficient evidence is not available to support the claimed health benefit and safe use of such oils. To address potential quality issues and identify possible adulteration, authenticated avocado oils extracted from the fruit peel, pulp and seed by supercritical fluid extraction (SFE), as well as commercial avocado pulp and seed oils sold in US market were analyzed for TGs and FAs in the present study. Characterization and quantification of TGs were conducted using UHPLC/ESI-MS. Thirteen TGs containing saturated and unsaturated fatty acids in avocado oils were unambiguously identified. Compared to traditional analytical methods, which are based only on the relative areas of chromatographic peaks neglecting the differences in the relative response of individual TG, our method improved the quantification of TGs by using the reference standards whenever possible or the reference standards with the same equivalent carbon number (ECN). To verify the precision and accuracy of the UHPLC/ESI-MS method, the hydrolysis and transesterification products of avocado oil were analyzed for fatty acid methyl esters using a GC/MS method. The concentrations of individual FA were calculated, and the results agreed with the UHPLC/ESI-MS method. Although chemical profiles of avocado oils from pulp and peel are very similar, a significant difference was observed for the seed oil. Principal component analysis (PCA) based on TG and FA compositional data allowed correct identification of individual avocado oil and detection of possible adulteration.

Computer Vision Classification of Barley Flour Based on Spatial Pyramid Partition Ensemble

Imaging sensors are largely employed in the food processing industry for quality control. Flour from malting barley varieties is a valuable ingredient in the food industry, but its use is restricted due to quality aspects such as color variations and the presence of husk fragments. On the other hand, naked varieties present superior quality with better visual appearance and nutritional composition for human consumption. Computer Vision Systems (CVS) can provide an automatic and precise classification of samples, but identification of grain and flour characteristics require more specialized methods. In this paper, we propose CVS combined with the Spatial Pyramid Partition ensemble (SPPe) technique to distinguish between naked and malting types of twenty-two flour varieties using image features and machine learning. SPPe leverages the analysis of patterns from different spatial regions, providing more reliable classification. Support Vector Machine (SVM), k-Nearest Neighbors (k-NN), J48 decision tree, and Random Forest (RF) were compared for samples' classification. Machine learning algorithms embedded in the CVS were induced based on 55 image features. The results ranged from 75.00% (k-NN) to 100.00% (J48) accuracy, showing that sample assessment by CVS with SPPe was highly accurate, representing a potential technique for automatic barley flour classification.

Multispectral Approach for Identifying Invasive Plant Species Based on Flowering Phenology Characteristics

Invasive plant species (IPS) are the second biggest threat to biodiversity after habitat loss. Since the spatial extent of IPS is essential for managing the invaded ecosystem, the current study aims at identifying and mapping the aggressive IPS of Acacia salicina and Acacia saligna, to understand better the key factors influencing their distribution in the coastal plain of Israel. This goal was achieved by integrating airborne-derived hyperspectral imaging and multispectral earth observation for creating species distribution maps. Hyperspectral data, in conjunction with high spatial resolution species distribution maps, were used to train the multispectral images at the species level. We incorporated a series of statistical models to classify the IPS location and to recognize their distribution and density. We took advantage of the phenological flowering stages of Acacia trees, as obtained by the multispectral images, for the support vector machine classification procedure. The classification yielded an overall Kappa coefficient accuracy of 0.89. We studied the effect of various environmental and human factors on IPS density by using a random forest machine learning model, to understand the mechanisms underlying successful invasions, and to assess where IPS have a higher likelihood of occurring. This algorithm revealed that the high density of Acacia most closely related to elevation, temperature pattern, and distances from rivers, settlements, and roads. Our results demonstrate how the integration of remote-sensing data with different data sources can assist in determining IPS proliferation and provide detailed geographic information for conservation and management efforts to prevent their future spread.

Alternative data mining/machine learning methods for the analytical evaluation of food quality and authenticity - A review

In recent years, the variety and volume of data acquired by modern analytical instruments in order to conduct a better authentication of food has dramatically increased. Several pattern recognition tools have been developed to deal with the large volume and complexity of available trial data. The most widely used methods are principal component analysis (PCA), partial least squares-discriminant analysis (PLS-DA), soft independent modelling by class analogy (SIMCA), k-nearest neighbours (kNN), parallel factor analysis (PARAFAC), and multivariate curve resolution-alternating least squares (MCR-ALS). Nevertheless, there are alternative data treatment methods, such as support vector machine (SVM), classification and regression tree (CART) and random forest (RF), that show a great potential and more advantages compared to conventional ones. In this paper, we explain the background of these methods and review and discuss the reported studies in which these three methods have been applied in the area of food quality and authenticity. In addition, we clarify the technical terminology used in this particular area of research.

Tracing Geographical Origins of Teas Based on FT-NIR Spectroscopy: Introduction of Model Updating and Imbalanced Data Handling Approaches

This work presents a reliable approach to trace teas' geographical origins despite changes in teas caused by different harvest years. A total of 1447 tea samples collected from various areas in 2014 (660 samples) and 2015 (787 samples) were detected by FT-NIR. Seven classifiers trained on the 2014 dataset all succeeded to trace origins of samples collected in 2014; however, they all failed to predict origins for the 2015 samples due to different data distributions and imbalanced dataset. Three outlier detection based undersampling approaches-one-class SVM (OC-SVM), isolation forest and elliptic envelope-were then proposed; as a result, the highest macro average recall (MAR) for the 2015 dataset was improved from 56.86% to 73.95% (by SVM). A model updating approach was also applied, and the prediction MAR was significantly improved with increase in the updating rate. The best MAR (90.31%) was first achieved by the OC-SVM combined SVM classifier at a 50% rate.

Heterogeneous Distribution of Erucic Acid in Brassica napus Seeds

Brassica napus (B. napus) is the world's most widely grown temperate oilseed crop. Although breeding for human consumption has led to removal of erucic acid from refined canola oils, there is renewed interest in the industrial uses of erucic acid derived from B. napus, and there is a rich germplasm available for use. Here, low- and high-erucic acid accessions of B. napus seeds were examined for the distribution of erucic acid-containing lipids and the gene transcripts encoding the enzymes involved in pathways for its incorporation into triacylglycerols (TAGs) across the major tissues of the seeds. In general, the results indicate that a heterogeneous distribution of erucic acid across B. napus seed tissues was contributed by two isoforms (out of six) of FATTY ACYL COA ELONGASE (FAE1) and a combination of phospholipid:diacylglycerol acyltransferase (PDAT)- and diacylglycerol acyltransferase (DGAT)-mediated incorporation of erucic acid into TAGs in cotyledonary tissues. An absence of the expression of these two FAE1 isoforms accounted for the absence of erucic acid in the TAGs of the low-erucic accession.

Multifractal analysis application to the study of fat and its infiltration in Iberian ham: Influence of racial and feeding factors and type of slicing

This paper explores the multifractal features of different commercial designations of Iberian ham (acorn 100% Iberian ham, acorn Iberian ham, feed/pasture Iberian ham and feed Iberian ham). This study has been done by taking as input the fatty infiltration patterns obtained from digital image analysis of ham cuts comparing mechanic and manual slicing. The yielded results show the multifractal nature of fatty connective tissue in Iberian ham, only when knife cutting is applied, confirming the differences between the designations according to their genetics and feeding. Thus, the multifractal parameters presented in this work could be considered as additional information for checking Iberian ham quality by using non-destructive methods based on the combination of image analysis and predictive techniques. Meat industry can take advantage of these methods to evaluate meat products, especially when fat-connective tissue with complex pattern distribution is involved.

Linking ecosystem services and ecosystem health to ecological risk assessment: A case study of the Beijing-Tianjin-Hebei urban agglomeration

Scientists have paid attention to the evaluation of the risk of ecosystem service degradation under rapid urbanization; yet the performance of the existing frameworks could be improved for tackling the challenges in the evaluation. In this study, a framework combining ecosystem service with ecosystem health as an assessing endpoint of ecological risk assessment was established. The framework was applied to investigate the way in which urbanization influences the ecosystem risk of the Beijing-Tianjin-Hebei urban agglomeration. Firstly, the decrease ratio of ecosystem service was mainly distributed in the range from 0 to 15%; the mean value of ecosystem health decreased from 0.402 to 0.311 from 2000 to 2010. The number of assessment units exhibiting risk degree grade I (the lowest risk degree grade) decreased by 7.03%, while the number of assessment units exhibiting risk degree grade V (the highest risk degree grade) increased by 1.61% from 2000 to 2010. The ratio of artificial surface should be controlled below 70%, based on the fitting model and for the purpose of resilience management. Overall, the analytical framework can comprehensively evaluate the impacts of complex practices in land-use planning on ecosystems.

Application of near infrared spectroscopy for rapid determination the geographical regions and polysaccharides contents of Lentinula edodes

In this study, a calibration model based on Near-infrared spectroscopy (NIR) technique and chemometrics method was developed for rapid and non-destructive detecting the polysaccharide contents of lentinula edodes samples collected from different regions. The polysaccharide contents of these samples were firstly determined by standard phenol-sulphruic acid method. Then, NIR spectra of these samples were collected by using Fourier transform infrared spectrometry. Based on these experimental data, a random forest method was further used to distinguish the regions of these samples, with a classification accuracy of 96.6%. After that, a rapid, accurate, and quantitative model was established for predicting the polysaccharide contents of these samples. In the model establishing process, some signal pre-treatment methods were optimized, and the validation results with highest determination coefficient (R²) and low root mean square errors of prediction (RMSEP) were, 0.925 and 0.720, respectively. These results showed that combined NIR technique with chemometrics was an effective and green method for lentinula edodes quality control.

GIS-Based Random Forest Weight for Rainfall-Induced Landslide Susceptibility Assessment at a Humid Region in Southern China

Landslide susceptibility assessment is presently considered an effective tool for landslide warning and forecasting. Under the assessment procedure, a credible index weight can greatly increase the rationality of the assessment result. Using the Beijiang River Basin, China, as a case study, this paper proposes a new weight-determining method based on random forest (RF) and used the weighted linear combination (WLC) to evaluate the landslide susceptibility. The RF weight and eight indices were used to construct the assessment model. As a comparison, the entropy weight (EW) and weight determined by analytic hierarchy process (AHP) were also used, respectively, to demonstrate the rationality of the proposed weight-determining method. The results show that: (1) the average error rates of training and testing based on RF are 18.12% and 15.83%, respectively, suggesting that the RF model can be considered rational and credible; (2) RF ranks the indices elevation (EL), slope (SL), maximum one-day precipitation (M1DP) and distance to fault (DF) as the Top 4 most important of the eight indices, occupying 73.24% of the total, while the indices runoff coefficient (RC), normalized difference vegetation index (NDVI), shear resistance capacity (SRC) and available water capacity (AWC) are less consequential, with an index importance degree of only 26.76% of the total; and (3) the verification of landslide susceptibility indicates that the accuracy rate based on the RF weight reaches 75.41% but are only 59.02% and 72.13% for the other two weights (EW and AHP), respectively. This paper shows the potential to provide a new weight-determining method for landslide susceptibility assessment. Evaluation results are expected to provide a reference for landslide management, prevention and reduction in the studied basin.

Antidiabetic Activity and Chemical Composition of Sanbai Melon Seed Oil

Objectives

Many fruits and herbs had been used in Traditional Chinese Medicines for treating diabetes mellitus (DM); however, scientific and accurate evidences regarding their efficacy and possible mechanisms were largely unknown. Sanbai melon seed oil (SMSO) was used in folk medicine in treating DM, but there is no literature about these effects. The present study was aimed at confirming the treatment effects of SMSO in type 1 DM.

Methods

Diabetes was induced by single intraperitoneal injection of streptozotocin (STZ) at a dose of 65 mg/kg body weight. After diabetes induction, mice were treated with SMSO at dose of 1 g/kg, 2 g/kg, and 4 g/kg. Drugs were given by gavage administration once a day continuously for 28 days. At the end of treatment, several biochemical parameters and molecular mechanisms were determined by biochemical assays, ELISA, and Western blotting. The chemical compositions of SMSO were also tested.

Results

SMSO treatment significantly improved the symptoms of weight loss, polydipsia, reduced FBG level, increased plasma insulin levels, reduced plasma lipids levels, and protected islet injury. The results also showed that SMSO mitigated oxidative stress and alleviated the liver and renal injury in diabetes mice. SMSO also protected islet cells from apoptotic damage by suppressing ER mediated and mitochondrial dependent apoptotic pathways. Further constituent analysis results showed that SMSO had rich natural resources which had beneficial effects on DM.

Conclusions

This study showed that SMSO had excellent antidiabetes effect and provided scientific basis for the use of SMSO as the functional ingredients production and dietary supplements production in the food and pharmaceutical industries.

Determination of 2-Propenal Using Headspace Solid-Phase Microextraction Coupled to Gas Chromatography–Time-of-Flight Mass Spectrometry as a Marker for Authentication of Unrefined Sesame Oil

Ascertaining the authenticity of the unrefined sesame oil presents an ongoing challenge. Here, the determination of 2-propenal was performed by headspace solid-phase microextraction (HS-SPME) under mild temperature coupled to gas chromatography with time-of-flight mass spectrometry, enabling the detection of adulteration of unrefined sesame oil with refined corn or soybean oil. Employing this coupled technique, 2-propenal was detected in all tested refined corn and soybean oils but not in any of the tested unrefined sesame oil samples. Using response surface methodology, the optimum extraction temperature, equilibrium time, and extraction time for the HS-SPME analysis of 2-propenal using carboxen/polydimethylsiloxane fiber were determined to be 55°C, 15 min, and 15 min, respectively, for refined corn oil and 55°C, 25 min, and 15 min, respectively, for refined soybean oil. Under these optimized conditions, the adulteration of unrefined sesame oil with refined corn or soybean oils (1–5%) was successfully detected. The detection and quantification limits of 2-propenal were found to be in the range of 0.008–0.010 and 0.023–0.031 µg mL−1, respectively. The overall results demonstrate the potential of this novel method for the authentication of unrefined sesame oil.

Proof-of-Principle of rTLC, an Open-Source Software Developed for Image Evaluation and Multivariate Analysis of Planar Chromatograms

High-performance thin-layer chromatography (HPTLC) is an advantageous analytical technique for analysis of complex samples. Combined with multivariate data analysis, it turns out to be a powerful tool for profiling of many samples in parallel. So far, chromatogram analysis has been time-consuming and required the application of at least two software packages to convert HPTLC chromatograms into a numerical data matrix. Hence, this study aimed to develop a powerful, all in one open-source software for user-friendly image processing and multivariate analysis of HPTLC chromatograms. Using the caret package for machine learning, the software was set up in the R programming language with an HTML-user interface created by the shiny package. The newly developed software, called rTLC, is deployed online, and instructions for direct use as a web application and for local installation, if required, are available on GitHub. rTLC was created especially for routine use in planar chromatography. It provides the necessary tools to guide the user in a fast protocol to the statistical data output (e.g., data extraction, preprocessing techniques, variable selection, and data analysis). rTLC offers a standardized procedure and informative visualization tools that allow the user to explore the data in a reproducible and comprehensive way. As proof-of-principle of rTLC, German propolis samples were analyzed using pattern recognition techniques, principal component analysis, hierarchic cluster analysis, and predictive techniques, such as random forest and support vector machines.

An Untargeted Metabolomics Approach to Characterize Short-Term and Long-Term Metabolic Changes after Bariatric Surgery

Bariatric surgery is currently one of the most effective treatments for obesity and leads to significant weight reduction, improved cardiovascular risk factors and overall survival in treated patients. To date, most studies focused on short-term effects of bariatric surgery on the metabolic profile and found high variation in the individual responses to surgery. The aim of this study was to identify relevant metabolic changes not only shortly after bariatric surgery (Roux-en-Y gastric bypass) but also up to one year after the intervention by using untargeted metabolomics. 132 serum samples taken from 44 patients before surgery, after hospital discharge (1-3 weeks after surgery) and at a 1-year follow-up during a prospective study (NCT01271062) performed at two study centers (Austria and Switzerland). The samples included 24 patients with type 2 diabetes at baseline, thereof 9 with diabetes remission after one year. The samples were analyzed by using liquid chromatography coupled to high resolution mass spectrometry (LC-HRMS, HILIC-QExactive). Raw data was processed with XCMS and drift-corrected through quantile regression based on quality controls. 177 relevant metabolic features were selected through Random Forests and univariate testing and 36 metabolites were identified. Identified metabolites included trimethylamine-N-oxide, alanine, phenylalanine and indoxyl-sulfate which are known markers for cardiovascular risk. In addition we found a significant decrease in alanine after one year in the group of patients with diabetes remission relative to non-remission. Our analysis highlights the importance of assessing multiple points in time in subjects undergoing bariatric surgery to enable the identification of biomarkers for treatment response, cardiovascular benefit and diabetes remission. Key-findings include different trend pattern over time for various metabolites and demonstrated that short term changes should not necessarily be used to identify important long term effects of bariatric surgery.