Comparison of hyperspectral imaging and spectrometers for prediction of cheeses composition.

Artisanal cheeses are part of the heritage and identity of different countries or regions. In this work, we investigated the spectral variability of a wide range of traditional Brazilian cheeses and compared the performance of different spectrometers to discriminate cheese types and predict compositional parameters. Spectra in the visible (vis) and near infrared (NIR) region were collected, using imaging (vis/NIR-HSI and NIR-HSI) and conventional (NIRS) spectrometers, and it was determined the chemical composition of seven types of cheeses produced in Brazil. Principal component analysis (PCA) showed that spectral variability in the vis/NIR spectrum is related to differences in color (yellowness index) and fat content, while in NIR there is a greater influence of productive steps and fat content. Partial least squares discriminant analysis (PLSDA) models based on spectral information showed greater accuracy than the model based on chemical composition to discriminate types of traditional Brazilian cheeses. Partial least squares (PLS) regression models based on vis/NIR-HSI, NIRS, NIR-HSI data and HSI spectroscopic data fusion (vis/NIR + NIR) demonstrated excellent performance to predict moisture content (RPD > 2.5), good ability to predict fat content (2.0 < RPD < 2.5) and can be used to discriminate between high and low protein values (∼1.5 < RPD < 2.0). The results obtained for imaging and conventional equipment are comparable and sufficiently accurate, so that both can be adapted to predict the chemical composition of the Brazilian traditional cheeses used in this study according to the needs of the industry.

Application of a microfluidic electronic tongue based on impedance spectroscopy for coconut water analysis.

The food industry has grown with the demands for new products and their authentication, which has not been accompanied by the area of analysis and quality control, thus requiring novel process analytical technologies for food processes. An electronic tongue (e-tongue) is a multisensor system that can characterize complex liquids in a fast and simple way. Here, we tested the efficacy of an impedimetric microfluidic e-tongue setup - comprised by four interdigitated electrodes (IDE) on a printed circuit board (PCB), with four pairs of digits each, being one bare sensor and three coated with different ultrathin nanostructured films with different electrical properties - in the analysis of fresh and industrialized coconut water. Principal Component Analysis (PCA) was applied to observe sample differences, and Partial Least Squares Regression (PLSR) was used to predict sample physicochemical parameters. Linear Discriminant Analysis (LDA) and Partial Least Square - Discriminant Analysis (PLS-DA) were compared to classify samples based on data from the e-tongue device. Results indicate the potential application of the microfluidic e-tongue in the identification of coconut water composition and determination of physicochemical attributes, allowing for classification of samples according to soluble solid content (SSC) and total titratable acidity (TTA) with over 90% accuracy. It was also demonstrated that the microfluidic setup has potential application in the food industry for quality assessment of complex liquid samples.

Characterization and classification of oleogels and edible oil using vibrational spectroscopy in tandem with one-class and multiclass chemometric methods.

Oleogel represents a promising healthier alternative to act as a substitute for conventional fat in various food products. Oil selection is a crucial factor in determining the technological properties and applications of oleogels due to their distinct fatty acid composition, molecular weight, and thermal properties, as well as the presence of antioxidants and oxidative stability. Hence, the relevance of monitoring oleogel properties by non-destructive, eco-friendly, portable, fast, and effective techniques is a relevant task and constitutes an advance in the evaluation of oleogels quality. Thus, the present study aims to classify oleogels rapidly and reliably, without the use of chemicals, comparing two handheld near infrared (NIR) spectrometers and one portable Raman device. Furthermore, two different multivariate methods are compared for oleogel classification according to oil type. Three types of oleogels were prepared, containing 95 % oil (sunflower, soy, olive) and 5 % beeswax as a structuring agent, melted at 90 °C. Polarized light microscopy (PLM) images were acquired, and fatty acid composition, peroxide index and free fatty acid content were determined using official methods. A total of 240 oleogel and 92 oil spectra were obtained for each instrument. After spectra pretreatment, Principal Component Analysis (PCA) was performed, and two classification methods were investigated. The Data Driven - Soft Independent Modelling of Class Analogy (DD-SIMCA) and Partial Least Squares Discriminant Analysis (PLS-DA) models demonstrated 95 % to 100 % of accuracy for the external test set. In conclusion, the use of vibrational spectroscopy using handheld and portable instruments in tandem with chemometrics showed to be an efficient alternative for classifying oils and oleogels and could be extended to other food samples. Although the classification of vegetable oils by NIR is widely used and known, this work proposes the classification of different types of oil in oleogel matrices, which has not yet been explored in the literature.

Correction: Low-cost electronic-nose (LC-e-nose) systems for the evaluation of plantation and fruit crops: recent advances and future trends.

Correction for 'Low-cost electronic-nose (LC-e-nose) systems for the evaluation of plantation and fruit crops: recent advances and future trends' by Marcus Vinicius da Silva Ferreira et al., Anal. Methods, 2023, https://doi.org/10.1039/D3AY01192E.

The Application of Handheld Near-Infrared Spectroscopy and Raman Spectroscopic Imaging for the Identification and Quality Control of Food Products.

The following investigations describe the potential of handheld NIR spectroscopy and Raman imaging measurements for the identification and authentication of food products. On the one hand, during the last decade, handheld NIR spectroscopy has made the greatest progress among vibrational spectroscopic methods in terms of miniaturization and price/performance ratio, and on the other hand, the Raman spectroscopic imaging method can achieve the best lateral resolution when examining the heterogeneous composition of samples. The utilization of both methods is further enhanced via the combination with chemometric evaluation methods with respect to the detection, identification, and discrimination of illegal counterfeiting of food products. To demonstrate the solution to practical problems with these two spectroscopic techniques, the results of our recent investigations obtained for various industrial processes and customer-relevant product examples have been discussed in this article. Specifically, the monitoring of food extraction processes (e.g., ethanol extraction of clove and water extraction of wolfberry) and the identification of food quality (e.g., differentiation of cocoa nibs and cocoa beans) via handheld NIR spectroscopy, and the detection and quantification of adulterations in powdered dairy products via Raman imaging were outlined in some detail. Although the present work only demonstrates exemplary product and process examples, the applications provide a balanced overview of materials with different physical properties and manufacturing processes in order to be able to derive modified applications for other products or production processes.

Low-cost electronic-nose (LC-e-nose) systems for the evaluation of plantation and fruit crops: recent advances and future trends.

An electronic nose (e-nose) is a device designed to recognize and classify odors. The equipment is built around a series of sensors that detect the presence of odors, especially volatile organic compounds (VOCs), and generate an electric signal (voltage), known as e-nose data, which contains chemical information. In the food business, the use of e-noses for analyses and quality control of fruits and plantation crops has increased in recent years. Their use is particularly relevant due to the lack of non-invasive and inexpensive methods to detect VOCs in crops. However, the majority of reports in the literature involve commercial e-noses, with only a few studies addressing low-cost e-nose (LC-e-nose) devices or providing a data-oriented description to assist researchers in choosing their setup and appropriate statistical methods to analyze crop data. Therefore, the objective of this study is to discuss the hardware of the two most common e-nose sensors: electrochemical (EC) sensors and metal oxide sensors (MOSs), as well as a critical review of the literature reporting MOS-based low-cost e-nose devices used for investigating plantations and fruit crops, including the main features of such devices. Miniaturization of equipment from lab-scale to portable and convenient gear, allowing producers to take it into the field, as shown in many appraised systems, is one of the future advancements in this area. By utilizing the low-cost designs provided in this review, researchers can develop their own devices based on practical demands such as quality control and compare results with those reported in the literature. Overall, this review thoroughly discusses the applications of low-cost e-noses based on MOSs for fruits, tea, and coffee, as well as the key features of their equipment (i.e., advantages and disadvantages) based on their technical parameters (i.e., electronic and physical parts). As a final remark, LC-e-nose technology deserves significant attention as it has the potential to be a valuable quality control tool for emerging countries.

Oxidation of whey protein isolate after thermal convection and microwave heating and freeze-drying: Correlation among physicochemical and NIR spectroscopy analyses.

This study investigated the oxidative susceptibility of whey protein isolate (WPI) dispersions treated by microwave or thermal convection before freeze-drying. WPI (20 mg protein/mL) in distilled water (DW) was heated at 63 ± 2 °C for 30 min by microwave (WPI-MW) or convection heating (WPI-CH) and freeze-dried. Untreated WPI (WPI-C), WPI solubilized in DW and freeze-dried (WPI-FD), and WPI solubilized in DW, heated at 98 ± 2 °C for 2 min and freeze-dried (WPI-B) were also evaluated. Structural changes (turbidity, ζ potential, SDS-PAGE, and near-infrared spectroscopy (NIR)) and protein oxidation (dityrosine, protein carbonylation, and SH groups) were investigated. WPI-FD showed alterations compared to WPI-C, mainly concerning carbonyl groups. Microwave heating increased carbonyl groups and dityrosine formation compared to conventional heating. NIR spectrum indicated changes related to the formation of carbonyl groups and PCA analysis allowed us to distinguish the samples according to carbonyl group content. The results suggest that NIR may contribute to monitoring oxidative changes in proteins resulting from processing.

Portable near-infrared (NIR) spectrometer and chemometrics for rapid identification of butter cheese adulteration.

Artisanal cheeses are highly valued around the world for their distinct sensory characteristics, thus being prone to adulteration by substituting authentic material for cheaper products, such as vegetable oil. In this work, we developed a method based on a portable NIR spectrometer as a non-destructive and low-cost alternative to identify adulteration in butter cheese. Dataset consisted of authentic and intentionally adulterated cheeses in the laboratory and commercial cheeses, which were identified as authentic and adulterated with vegetable oil after analysis of the fatty acid profile. PLS-DA classification models identified adulterated samples with an accuracy of 94.44%. PLS prediction models showed excellent performance (RPD > 3.0) to predict the adulterant level. These results demonstrate that NIR spectra can be used to identify the replacement of authentic fat by soybean oil in butter cheese and that the developed models can be used to identify adulteration in external samples with good performance.

Rapid and non-destructive cinnamon authentication by NIR-hyperspectral imaging and classification chemometrics tools.

Cinnamon is a valuable aromatic spice widely used in pharmaceutical and food industry. Commonly, two-cinnamon species are available in the market, Cinnamomum verum (true cinnamon), cropped only in Sri Lanka, and Cinnamomum cassia (false cinnamon), cropped in different geographical origins. Thus, this work aimed to develop classification models based on NIR-hyperspectral imaging (NIR-HSI) coupled to chemometrics to classify C. verum and C. cassia sticks. First, principal component analysis (PCA) was applied to explore hyperspectral images. Scores surface displayed the high similarity between species supported by comparable macronutrient concentration. PC3 allowed better class differentiation compared to PC1 and PC2, with loadings exhibiting peaks related to phenolics/aromatics compounds, such as coumarin (C. cassia) or catechin (C. verum). Partial least square discriminant analysis (PLS-DA) and Support vector machine (SVM) reached similar performance to classify samples according to origin, with error = 3.3 % and accuracy = 96.7 %. A permutation test with p < 0.05 validated PLS-DA predictions have real spectral data dependency, and they are not result of chance. Pixel-wise (approach A) and sample-wise (approach B, C and D) classification maps reached a correct classification rate (CCR) of 98.3 % for C. verum and 100 % for C. cassia. NIR-HSI supported by classification chemometrics tools can be used as reliable analytical method for cinnamon authentication.

Data reduction by randomization subsampling for the study of large hyperspectral datasets.

Large amount of information in hyperspectral images (HSI) generally makes their analysis (e.g., principal component analysis, PCA) time consuming and often requires a lot of random access memory (RAM) and high computing power. This is particularly problematic for analysis of large images, containing millions of pixels, which can be created by augmenting series of single images (e.g., in time series analysis). This tutorial explores how data reduction can be used to analyze time series hyperspectral images much faster without losing crucial analytical information. Two of the most common data reduction methods have been chosen from the recent research. The first one uses a simple randomization method called randomized sub-sampling PCA (RSPCA). The second implies a more robust randomization method based on local-rank approximations (rPCA). This manuscript exposes the major benefits and drawbacks of both methods with the spirit of being as didactical as possible for a reader. A comprehensive comparison is made considering the amount of information retained by the PCA models at different compression degrees and the performance time. Extrapolation is also made to the case where the effect of time and any other factor are to be studied simultaneously.

Near-infrared techniques for fraud detection in dairy products: A review.

The dairy products sector is an important part of the food industry, and their consumption is expected to grow in the next 10 years. Therefore, the authentication of these products in a faster and precise way is required for the sake of public health. This review proposes the use of near-infrared techniques for the detection of food fraud in dairy products as they are faster, nondestructive, environmentally friendly, do not require sample preparation, and allow multiconstituent analysis. First, we have described frequent forms of food fraud in dairy products and the application of traditional techniques for their detection, highlighting gaps and counterproductive characteristics for the actual global food chain, as longer sample preparation time and use of reagents. Then, the application of near-infrared spectroscopy and hyperspectral imaging for the detection of food fraud mainly in cheese, butter, and yogurt are described. As these techniques depend on model development, the coverage of different dairy products by the literature will promote the identification of food fraud in a faster and reliable way.

Incompatibility between sodium caseinate - locust bean gum induced by NaCl and yerba mate extract.

Aqueous two-phase system (ATPS) is a technique used for the separation of biopolymers in two aqueous phases. Some combinations of biopolymers can form a water-in-water (W/W) emulsion due to steric exclusion and thermodynamic incompatibility between these biopolymers under some specific conditions. In this work, the formation of W/W emulsions composed of sodium caseinate (SCN) and locust bean gum (LBG) was evaluated, using NaCl or yerba mate extract as the driving force for the phase separation, which was described by phase's diagrams. Phase diagrams are like fingerprints of ATPS systems, which demonstrate the specific conditions to develop separate phases. Phase diagrams of the two systems show that at the same concentrations of protein and carbohydrate, the addition of NaCl or extract induced the separation of the compounds differently. Salt promotes phase separation by steric exclusion, each phase being rich in one of the polymers. Since extract may also induce other effects, such as the formation of a SCN-extract-LBG complex, migration of LBG to the SCN-rich phase was promoted, modifying the characteristics of the tie lines in the phase diagrams. However, it was feasible to separate the protein in systems containing concentrated phenolic extract, whose incorporation is relevant considering its antioxidant activity.

Near infrared hyperspectral imaging and spectral unmixing methods for evaluation of fiber distribution in enriched pasta.

Pasta is mostly composed by wheat flour and water. Nevertheless, flour can be partially replaced by fibers to provide extra nutrients in the diet. However, fiber can affect the technological quality of pasta if not properly distributed. Usually, determinations of parameters in pasta are destructive and time-consuming. The use of Near Infrared-Hyperspectral Imaging (NIR-HSI), together with machine learning methods, is valuable to improve the efficiency in the assessment of pasta quality. This work aimed to investigate the ability of NIR-HSI and augmented Multivariate Curve Resolution-Alternating Least Squares (MCR-ALS) for the evaluation, resolution and quantification of fiber distribution in enriched pasta. Results showed R2V between 0.28 and 0.89, %LOF < 6%, variance explained over 99%, and similarity between pure and recovered spectra over 96% and 98% in models using pure flour and control as initial estimates, respectively, demonstrating the applicability of NIR-HSI and MCR-ALS in the identification of fiber in pasta.

Occurrence of wooden breast and white striping in Brazilian slaughtering plants and use of near-infrared spectroscopy and multivariate analysis to identify affected chicken breasts.

White Striping (WS) and Wooden Breast (WB) are emerging poultry myopathies that occur worldwide, affecting the quality of meat. The aim of this study was to evaluate the occurrence of N, WS, WB, and WS/WB (myopathies combined) in chicken breast from Brazilian commercial plant, comparing (1) inspection based on visual aspect and palpation of Pectoralis major muscle, and (2) identification of these myopathies by near-infrared Spectroscopy (NIRS). Chickens slaughtered at Brazilian commercial plant at four age ranges (4 to 5, 6 to 7, 8 to 9, and 65 weeks) were inspected. Spectral information was acquired using a portable NIR spectrometer, and classification models were performed using and Successive Projection Algorithm-Linear Discriminant Analysis (SPA-LDA) and Soft Independent Modeling of Class Analogy (SIMCA) to distinguish normal and affected muscles. Results showed that occurrence of myopathies was aggravated by age of slaughter, as chicken slaughtered at 4 to 5 and 65 weeks exhibited 13.6 and 95% of myopathies, respectively. Birds slaughtered at 65 weeks showed no occurrence of WB, isolated or combined with WS. It was not possible to differentiate the WB and WS/WB classes; therefore, those samples were grouped (WB+WS/WB). SPA-LDA model showed greater accuracy (92 to 93%) in identifying Normal (N), WS, and WB+WS/WB groups, compared to SIMCA (89 to 91%). It can be concluded that the level of occurrence of myopathies in meat is directly related to the age of slaughter. This study demonstrated that NIRS combined with SPA-LDA model could be used as a tool to detect myopathies in chicken breast. This technique has potential for application in industrial processing lines as an alternative to the traditional methods of identification. PRACTICAL APPLICATION: This study shows that NIRS combined with chemometric techniques can be used to identify chicken breast myopathies in a wide range of ages at slaughter. In addition to being able to discriminate chicken muscles into subclasses, namely, Normal, WS, and WB/WB+WS, this technique has potential for application in industrial processing lines as it is a portable and nondestructive method. This procedure is emphasized as an alternative to the conventional method of identification based on palpation and visual assessment of muscle.

Metal Accumulation by Jatropha curcas L. Adult Plants Grown on Heavy Metal-Contaminated Soil.

Jatropha curcas has the ability to phytoextract high amounts of heavy metals during its first months just after seeding. Notwithstanding, there is scarce information about metal uptake by adult J. curcas plants. To shed light on this issue, 4-year-old J. curcas L. plants were planted in a soil mixture of peat moss and mining soil (high metals content), and the biomass growth and metal absorption during 90 days were compared with those of plants growing in peat moss. The main metal found in the mining soil was Fe (31985 mg kg-1) along with high amounts of As (23717 mg kg-1). After the 90-day phytoremediation, the plant removed 29% of Fe and 44% of As from the soil mixture. Results revealed that J. curcas L. translocated high amounts of metals to its aerial parts, so that translocation factors were much higher than 1. Because of the high translocation and bioaccumulation factors obtained, J. curcas L. can be regarded as a hyperaccumulator plant. Despite the great capacity of J. curcas L. to phytoremediate heavy-metal-contaminated soils, the main drawback is the subsequent handling of the metal-contaminated biomass, although some potential applications have been recently highlighted for this biomass.

Determination of pectin content in orange peels by near infrared hyperspectral imaging.

Pectin has several purposes in the food and pharmaceutical industry making its quantification important for further extraction. Current techniques for pectin quantification require its extraction using chemicals and producing residues. Determination of pectin content in orange peels was investigated using near infrared hyperspectral imaging (NIR-HSI). Hyperspectral images from orange peel (140 samples) with different amounts of pectin were acquired in the range of 900-2500 nm, and the spectra was used for calibration models using multivariate statistical analyses. Principal component analysis (PCA) and linear discriminant analysis (LDA) showed better results considering three groups: low (0-5%), intermediate (10-40%) and high (50-100%) pectin content. Partial least squares regression (PLSR) models based on full spectra showed higher precision (R2 > 0.93) than those based on few selected wavelengths (R2 between 0.92 and 0.94). The results demonstrate the potential of NIR-HSI to quantify pectin content in orange peels, providing a valuable technique for orange producers and processing industries.

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.

Identification of fiber added to semolina by near infrared (NIR) spectral techniques.

Ingredients added in food products can increase the nutritional value, but also affect their functional properties. After processing, determination of added ingredients is difficult, thus it is important to develop rapid techniques for quantification of food ingredients. In the current work, near infrared spectroscopy (NIRS) and hyperspectral imaging (NIR-HSI) were investigated to quantify the amount of fiber added to semolina and its distribution. NIR spectra were acquired to compare the accuracy in the classification, quantification and distribution of fibers added to semolina. Principal Component Analyses (PCA) and Soft Independent Modeling of Class Analogy (SIMCA) were used for classification. Partial Least Squares Regression (PLSR) models applied to NIR-HSI spectra showed R2P between 0.85 and 0.98, and RMSEP between 0.5 and 1%, and were used for prediction map of the samples. These results showed that NIR-HSI technique can be used for the identification and quantification of fiber added to semolina.

Portable NIR Spectrometer for Prediction of Palm Oil Acidity.

Palm oil is widely used in the food industry, and its quality is associated with the free fatty acids (FFA) content. Determination of FFA in oil is time-consuming, requires chemicals and generates residues. There is a trend of applying process analytical technologies (PAT) for fast and nondestructive determination of oil parameters. Portable near-infrared (NIR) spectrometers are cheaper than bench top equipment, and have been used for several tasks in the food processing industry, as it provides fast and reliable data for inline measurements. This study investigated the use of NIR spectra using a portable equipment, combined with both unsupervised and supervised multivariate analyses for identification of palm oil samples with different levels of FFA. Soft independent modeling of class analogy , k-Nearest Neighbors, and linear discriminant analysis models were able to correctly identify 100% of the studied samples with selected wavelengths from NIR spectra. Calibration models were performed for acidity prediction, achieving R2 = 0.97, with root mean square error of prediction = 4.37 for partial least squares model using most relevant wavelengths. These results demonstrate the feasibility of applying a low-cost portable NIR spectrophotometer to predict quality parameters of palm oil. PRACTICAL APPLICATION: This work presents results that show the feasibility of using a low-cost portable near-infrared spectrophotometer for the classification of raw palm oil samples according to free fatty acids contents. Regression models are presented as a fast and nondestructive alternative to classify samples for acidity, which is an important quality parameter and that directly affects the market value of crude palm oil.

Nontargeted Analytical Methods as a Powerful Tool for the Authentication of Spices and Herbs: A Review.

Food fraud in herbs and spices is an important topic, which has led to new technologies being studied as potential tools for fraud identification. Nontargeted technologies have proven to be a useful tool for the authentication of herbs and spices. The present review focuses on the use of near-infrared, hyperspectral imaging, Fourier-transform infrared, Raman, nuclear magnetic resonance, and electron spin resonance spectroscopy for the authentication of spices, which includes the determination of origin and irradiated spices and the identification of adulterants. The methods developed based on vibrational spectroscopy combined with chemometric techniques seem to be promising tools for determining the presence of adulterants and contaminants in herbs and spices. On the other hand, nuclear magnetic resonance seems to be the most efficient technology to determine the origin of herbs and spices although, for some cases, studies with near-infrared spectroscopy can be a viable substitute. Electron spin resonance spectroscopy is the technique par excellence used for the authentication of irradiated herbs and spices, so its use should be expanded to many more spices' varieties. Portable devices are preferred by those involved in the food industry, due to its manageability and low cost. Data fusion and big data are shown as promising tools for spice fraud control. In conclusion, spectroscopic techniques show a great efficiency to authenticate spices, although their evaluation must be expanded to other spice varieties, to new strategies of data analysis (as data fusion and big data), and to the use of portable devices.