In situ prediction of phenolic compounds in puff dried Ziziphus jujuba Mill. using hand-held spectral analytical system

Improvement of near-infrared spectroscopic assessment methods for the quality of Keemun black tea: Utilizing transfer learning

Keemun black tea, a renowned Chinese black tea, presents challenges in quality assessment due to variability in data across different years. To address this, we developed transfer learning algorithms using near-infrared spectral data. The qualitative algorithm, BLACK TEA-GRADING, utilizes stochastic Fourier features of the limit learning machine, improving F1 scores from 0.7035 to 0.8138 in grading tea across different years. The quantitative algorithm, BLACK TEA-SUBSTANCE ANALYSIS, which is based on the inverse Gram matrix, was developed to predict the main flavor substances. Without the algorithm, the MSE exceeded 0.04 and R values were below 0.7; with it, MSE dropped below 0.015 and R values exceeded 0.8. Therefore, our method leverages transfer learning and near-infrared spectroscopy to enhance the accuracy of Keemun black tea quality assessment across different years, promoting the use of near-infrared spectroscopy in tea quality evaluation.

Comparison of Multiple NIR Instruments for the Quantitative Evaluation of Grape Seed and Other Polyphenolic Extracts with High Chemical Similarities

Grape seed extract (GSE), one of the world's bestselling dietary supplements, is prone to frequent adulteration with chemically similar compounds. These frauds can go unnoticed within the supply chain due to the use of unspecific standard analytical methods for quality control. This research aims to develop a near-infrared spectroscopy (NIRS) method for the rapid and non-destructive quantitative evaluation of GSE powder in the presence of multiple additives. Samples were prepared by mixing GSE with pine bark extract (PBE) and green tea extract (GTE) on different levels between 0.5 and 13% in singular and dual combinations. Measurements were performed with a desktop and three different handheld devices for performance comparison. Following spectral pretreatment, partial least squares regression (PLSR) and support vector regression (SVR)-based quantitative models were built to predict extract concentrations and various chemical parameters. Cross- and external-validated models could reach a minimum R2p value of 0.99 and maximum RMSEP of 0.27% for the prediction of extract concentrations using benchtop data, while models based on handheld data could reach comparably good results, especially for GTE, caffeic acid and procyanidin content prediction. This research shows the potential applicability of NIRS coupled with chemometrics as an alternate, rapid and accurate quality evaluation tool for GSE-based supplement mixtures.

Rapid and comprehensive grade evaluation of Keemun black tea using efficient multidimensional data fusion

To develop a comprehensive evaluation method for Keemun black tea, we used micro-near-infrared spectroscopy, computer vision, and colorimetric sensor array to collect data. We used support vector machine, least-squares support vector machine (LS-SVM), extreme learning machine, and partial least squares discriminant analysis algorithms to qualitatively discriminate between different grades of tea. Our results indicated that the LS-SVM model with mid-level data fusion attained an accuracy of 98.57% in the testing set. To quantitatively determine flavour substances in black tea, we used support vector regression. The correlation coefficient for the predicted sets of gallic acid, caffeine, epigallocatechin, catechin, epigallocatechin gallate, epicatechin, gallocatechin gallate and total catechins were 0.84089, 0.94249, 0.94050, 0.83820, 0.81111, 0.82670, 0.93230, and 0.93608, respectively. Furthermore, all compounds exhibited residual predictive deviation values exceeding 2. Hence, combining spectral, shape, colour, and aroma data with mid-level data can provide a rapid and comprehensive assessment of Keemun black tea quality.

Comparison of Si-GA-PLS and Si-CARS-PLS build algorithms for quantitation of total polyphenols in black tea using the spectral analytical system

BACKGROUND

The objective of the current study was to compare two machine learning approaches for the quantification of total polyphenols by choosing the optimal spectral intervals utilizing the synergy interval partial least squares (Si-PLS) model. To increase the resilience of built models, the genetic algorithm (GA) and competitive adaptive reweighted sampling (CARS) were applied to a subset of variables.

RESULTS

The collected spectral data were divided into 19 sub-interval selections totaling 246 variables, yielding the lowest root mean square error of cross-validation (RMSECV). The performance of the model was evaluated using the correlation coefficient for calibration (RC ), prediction (RP ), RMSECV, root mean square error of prediction (RMSEP) and residual predictive deviation (RPD) value. The Si-GA-PLS model produced the following results: PCs = 9; RC  = 0.915; RMSECV = 1.39; RP  = 0.8878; RMSEP = 1.62; and RPD = 2.32. The performance of the Si-CARS-PLS model was noted to be best at PCs = 10, while RC  = 0.9723, RMSECV = 0.81, RP  = 0.9114, RMSEP = 1.45 and RPD = 2.59.

CONCLUSION

The build model's prediction ability was amended in the order PLS < Si-PLS < CARS-PLS when full spectroscopic data were used and Si-PLS < Si-GA-PLS < Si-CARS-PLS when interval selection was performed with the Si-PLS model. Finally, the developed method was successfully used to quantify total polyphenols in tea. © 2023 Society of Chemical Industry.

Fusion-based strategy of CSA and mobile NIR for the quantification of free fatty acid in wheat varieties coupled with chemometrics

The use of sensor fusion, a novel method of combining artificial senses, has become increasingly popular in the assessment of food quality. This study employed a combination of the colorimetric sensor array (CSA) and mobile near-infrared (NIR) spectroscopy to predict free fatty acids in wheat flour. In conjunction with a partial least squares model, Low- and mid-level fusion strategies were used for quantification. Accordingly, performance of the built model was evaluated based on higher correlation coefficients between calibration and prediction (R_C and R_P), lower root mean square error of prediction (RMSEP), and a higher residual predictive deviation (RPD). The mid-level fusion coupled PLS model produced superior data fusion findings, with R_C = 0.8793, RMSECV = 7.91 mg/100 g, R_P = 0.8747, RMSEP = 6.99 mg/100 g, and RPD = 2.27. The findings of the study suggest that the NIR-CSA fusion approach could be effectively applied to the prediction of free fatty acids in wheat flour.

The Use of Infrared Spectroscopy for the Quantification of Bioactive Compounds in Food: A Review

Infrared spectroscopy (wavelengths ranging from 750-25,000 nm) offers a rapid means of assessing the chemical composition of a wide range of sample types, both for qualitative and quantitative analyses. Its use in the food industry has increased significantly over the past five decades and it is now an accepted analytical technique for the routine analysis of certain analytes. Furthermore, it is commonly used for routine screening and quality control purposes in numerous industry settings, albeit not typically for the analysis of bioactive compounds. Using the Scopus database, a systematic search of literature of the five years between 2016 and 2020 identified 45 studies using near-infrared and 17 studies using mid-infrared spectroscopy for the quantification of bioactive compounds in food products. The most common bioactive compounds assessed were polyphenols, anthocyanins, carotenoids and ascorbic acid. Numerous factors affect the accuracy of the developed model, including the analyte class and concentration, matrix type, instrument geometry, wavelength selection and spectral processing/pre-processing methods. Additionally, only a few studies were validated on independently sourced samples. Nevertheless, the results demonstrate some promise of infrared spectroscopy for the rapid estimation of a wide range of bioactive compounds in food matrices.

Simultaneous quantification of total flavonoids and phenolic content in raw peanut seeds via NIR spectroscopy coupled with integrated algorithms

Peanuts are nutritionally valuable for both humans and animals due to their high content of flavonoids and phenolic compounds. Herein, we explored the potential of near-infrared (NIR) spectroscopy coupled with efficient variable selection algorithms for quantitative prediction of total flavonoids (TFC) and total phenolics content (TPC) in raw peanut seeds. Spectrophotometrically, the reference results of the extracts for TFC and TPC were analysed and recorded. The integrated application of the synergy interval coupled competitive adaptive reweighted sampling-partial least squares (Si-CARS-PLS) were used for prediction. The model performance appraisal was based on the correlation coefficients of prediction (Rp), root mean square error of prediction (RMSEP), and residual predictive deviation (RPD). The Si-CARS-PLS performed optimally for TFC (Rp = 0.9137, RPD = 2.49) and TPC (Rp = 0.9042, RPD = 2.31), respectively. Moreover, the model (Si-CARS-PLS) was found to have an acceptable fit for the analytes under study since it achieved 0.88 for TFC and 0.86 for TPC based on the external validation. Therefore, these results showed that NIR coupled with Si-CARS-PLS could be used for the quantitative prediction of flavonoids and phenolic contents in raw peanut seeds.

Intelligent evaluation of free amino acid and crude protein content in raw peanut seed kernels using NIR spectroscopy paired with multivariable calibration

Given the nutritional importance of peanuts, this study examined the free amino acid (FAA) and crude protein (CP) content in raw peanut seeds. Near-infrared spectroscopy (NIRS) was employed in combination with variable selection algorithms after successful reference data analysis using colorimetric and Kjeldahl methods. Ensuing the application of partial least squares (PLS) as a full spectral model, the genetic algorithm (GA), bootstrapping soft shrinkage (BOSS), uninformative variable elimination (UVE), and random frog (RF) models were tested and assessed. A comparison of correlation coefficients of prediction (R_p), root mean square error of prediction (RMSEP), and residual predictive deviation (RPD) was performed to appraise the performance of the built models. Using RF-PLS, an unsurpassed outcome was achieved for FAA (R_p = 0.937, RPD = 3.38) and CP (R_p = 0.9261, RPD = 3.66). These findings demonstrated that NIR in combination with RF-PLS could be utilized for quantitative, rapid, and nondestructive prediction of FAA and CP in raw peanut seed samples.

A Narrative Review of Recent Advances in Rapid Assessment of Anthocyanins in Agricultural and Food Products

Anthocyanins (ACNs) are plant polyphenols that have received increased attention recently mainly due to their potential health benefits and applications as functional food ingredients. This has also created an interest in the development and validation of several non-destructive techniques of ACN assessments in several food samples. Non-destructive and conventional techniques play an important role in the assessment of ACNs in agricultural and food products. Although conventional methods appear to be more accurate and specific in their analysis, they are also associated with higher costs, the destruction of samples, time-consuming, and require specialized laboratory equipment. In this review article, we present the latest findings relating to the use of several spectroscopic techniques (fluorescence, Raman, Nuclear magnetic resonance spectroscopy, Fourier-transform infrared spectroscopy, and near-infrared spectroscopy), hyperspectral imaging, chemometric-based machine learning, and artificial intelligence applications for assessing the ACN content in agricultural and food products. Furthermore, we also propose technical and future advancements of the established techniques with the need for further developments and technique amalgamations.

Comparative Determination of Phenolic Compounds in Arabidopsis thaliana Leaf Powder under Distinct Stress Conditions Using Fourier-Transform Infrared (FT-IR) and Near-Infrared (FT-NIR) Spectroscopy

The increasing interest in plant phenolic compounds in the past few years has become necessary because of their several important physicochemical properties. Thus, their identification through non-destructive methods has become crucial. This study carried out comparative non-destructive measurements of Arabidopsis thaliana leaf powder sample phenolic compounds using Fourier-transform infrared and near-infrared spectroscopic techniques under six distinct stress conditions. The prediction analysis of 600 leaf powder samples under different stress conditions (LED lights and drought) was performed using PLSR, PCR, and NAS-based HLA/GO regression analysis methods. The results obtained through FT-NIR spectroscopy yielded the highest correlation coefficient (Rp2) value of 0.999, with a minimum error (RMSEP) value of 0.003 mg/g, based on the PLSR model using the MSC preprocessing method, which was slightly better than the correlation coefficient (Rp2) value of 0.980 with an error (RMSEP) value of 0.055 mg/g for FT-IR spectroscopy. Additionally, beta coefficient plots present spectral differences and the identification of important spectral signatures sensitive to the phenolic compounds in the measured powdered samples. Thus, the obtained results demonstrated that FT-NIR spectroscopy combined with partial least squares regression (PLSR) and suitable preprocessing method has a solid potential for non-destructively predicting phenolic compounds in Arabidopsis thaliana leaf powder samples.

Variable selection in the chemometric treatment of food data: A tutorial review

Food analysis covers aspects of quality and detection of possible frauds to ensure the integrity of the food. The arsenal of analytical instruments available for food analysis is broad and allows the generation of a large volume of information per sample. But this instrumental information may not yet give the desired answer; it must be processed to provide a final answer for decision making. The possibility of discarding non-informative and/or redundant signals can lead to models of better accuracy, robustness, and chemical interpretability, in line with the principle of parsimony. Thus, in this tutorial review, we cover aspects of variable selection in food analysis, including definitions, theoretical aspects of variable selection, and case studies showing the advantages of variable selection-based models concerning the use of a wide range of non-informative and redundant instrumental information in the analysis of food matrices.

Application of benchtop NIR spectroscopy coupled with multivariate analysis for rapid prediction of antioxidant properties of walnut (Juglans regia)

Benchtop near-infrared (NIR) spectroscopy coupled with multivariate analysis was used for the classification and prediction of antioxidant properties of walnut. Total phenolic content (TPC), total flavonoid content (TFC), ABTS assay and FRAP assay were performed spectrophotometrically. The synergy interval partial least square coupled competitive adaptive reweighted sampling (Si-CARS-PLS) was used for the prediction. A decent discrimination using principal component analysis (PCA) was observed by mean of spectroscopic and antioxidant properties data with total cumulative variance of 99.26% (PC1 = 95.07%, PC2 = 2.98%, PC3 = 1.21%) and 96.60% (PC1 = 64.28%, PC2 = 32.32%) respectively. The Si-CARS-PLS yielded optimal performance, R_P = 0.9616, RPD = 3.807 for TPC, R_P = 0.9657, RPD = 3.367 for TFC, R_P = 0.9683, RPD = 2.728 for ABTS assay, and R_P = 0.914, RPD = 2.669 for FRAP assay. These findings revealed that NIR integrated with Si-CARS-PLS could be used for the prediction of antioxidant properties of walnut.