Near Infrared Spectroscopy Spectroscopy Research Articles

Differentiation between fresh and frozen-thawed mackerel fish using low-cost portable near infrared spectrometry devices

Mislabelling frozen-thawed fish fillets as fresh is one of the most important fraudulent practices during commercialisation. This study aimed to determine the ability of two portable miniaturised low-cost near-infrared spectroscopy (NIR) devices intended for consumers to discriminate between fresh and thawed mackerel (submitted to one or two frozen-thawed cycles) and between different freezing systems. The effect of different fish seasonal characteristics on the performance of the model was also evaluated. The low-cost NIR devices were able to discriminate between fresh and thawed samples with a classification rate of 90.3% and 94.1% and the freezing system to which they were submitted with a classification rate of 91.2% and 89.7%. These findings suggest that low-cost portable NIR spectroscopy can be a valuable tool for detecting mislabeled frozen-thawed products sold as fresh, providing consumers with a rapid and affordable method for fraud detection.

Effect of Sample Presentation on the Classification of Black Soldier Fly Larvae Using Near-Infrared Spectroscopy

Black soldier fly larvae (BSFL) (Hermetia illucens) reared on food waste streams are considered a sustainable source of protein in feed livestock diets. Recently, portable near-infrared spectroscopy (NIR) instruments have been assessed to monitor the consistency and quality of food waste streams used to feed black soldier fly larvae. During the application of NIR spectroscopy, sample presentation (e.g., drying, processing, particle size) plays an important role in the accuracy of the models developed (quantitative or qualitative analysis). The objective of this study was to evaluate the effect of sample presentation (number of larvae used during the scanning of BSFL) on the accuracy of classification models developed to trace the food waste stream (e.g., supermarket of childcare) used to feed the larvae. BSFL samples were sourced from two waste streams and scanned as half, 1, 2, or 3 larvae using an NIR portable instrument (MicroNIR, Viavi, Milpitas, California, USA). Principal component analysis (PCA) and linear discriminant analysis (LDA) were used to analyze the NIR data and to classify the samples according to the waste stream. The main differences in the NIR spectra of the BSFL samples associated with the number of larvae scanned were observed around 1200 nm, mainly associated with the C-H overtones (lipids). The classification results showed that high classification rates (>93%) were obtained regardless of the number of larvae scanned, ranging from 93% (using 0.5 larvae) to 100% (using 1, 2, or 3 larvae samples). Overall, the number of larvae scanned had minimal to no effect on the accuracy of the LDA classification models. The present study demonstrated that a portable NIR instrument can be suitable for an initial rapid classification or determination of the origin of the waste stream used to feed the BSFL.

Quantitative analysis of three ingredients in Salvia miltiorrhiza by near infrared spectroscopy combined with hybrid variable selection strategy

Rosmarinic acid (RA), Tanshinone IIA (Tan IIA), and Salvianolic acid B (Sal B) are crucial compounds found in Salvia miltiorrhiza. Quickly predicting these components can aid in ensuring the quality of S. miltiorrhiza. Spectral preprocessing and variable selection are essential processes in quantitative analysis using near infrared spectroscopy (NIR). A novel hybrid variable selection approach utilizing iVISSA was employed in this study to enhance the quantitative measurement of RA, Tan IIA, and Sal B contents in S. miltiorrhiza. The spectra underwent 108 preprocessing approaches, with the optimal method being determined as orthogonal signal correction (OSC). iVISSA was utilized to identify the intervals (feature bands) that were most pertinent to the target chemical. Various methods such as bootstrapping soft shrinkage (BOSS), competitive adaptive reweighted sampling (CARS), genetic algorithm (GA), variable combination population analysis (VCPA), successive projections algorithm (SPA), iteratively variable subset optimization (IVSO), and iteratively retained informative variables (IRIV) were used to identify significant feature variables. PLSR models were created for comparison using the given variables. The results fully demonstrated that iVISSA-SPA calibration model had the best comprehensive performance for Tan IIA, and iVISSA-BOSS had the best comprehensive performance for RA and Sal B, and correlation coefficients of cross-validation (R2cv), root mean square errors of cross-validation (RMSECV), correlation coefficients of prediction (R2p), and root mean square errors of prediction (RMSEP) were 0.9970, 0.0054, 0.9990 and 0.0033, 0.9992, 0.0016, 0.9961 and 0.0034, 0.9998, 0.0138, 0.9875 and 0.1090, respectively. The results suggest that NIR spectroscopy, along with PLSR and a hybrid variable selection method using iVISSA, can be a valuable tool for quickly quantifying RA, Sal B, and Tan IIA in S. miltiorrhiza.

Enhancing quality control in emulsion-type sausage production: Predicting chemical composition of intact samples with near infrared spectroscopy

This research actively explores the potential of near infrared spectroscopy (NIR) for analyzing the chemical composition of emulsion-type sausages, focusing on critical factors like residual nitrite, moisture, protein, and fat content. To establish robust and generalizable models, we utilized a dataset of 100 experimentally prepared sausages encompassing a wide range of pork back fat replacement levels (5%, 15%, 30%, 45%, and 60%) and added sodium nitrite amounts (0, 80, 125, 250, and 375 ppm). An external validation set of 20 commercially sourced sausages further assessed the model’s real-world applicability. Partial least squares (PLS) regression calibration models with multiplicative scatter correction (MSC) pre-treatment demonstrated impressive accuracy for moisture (RMSECV = 0.57%, RPD = 9.8), fat (RMSECV = 1.17%, RPD = 9.5), and protein (RMSECV = 0.30%, RPD = 7.6). While residual nitrite prediction presented challenges due to its inherent complexity, the external validation yielded a competitive root mean square error of prediction (RMSEP) of 12.02 ppm, surpassing the average performance reported in similar studies (RMSEP ∼15 ppm) by 3 ppm. Importantly, sample homogenization did not significantly affect parameter prediction, highlighting the robustness of the NIR-based approach. These findings suggest that NIR spectroscopy, with its non-destructive, rapid, and cost-effective nature, could provide valuable tools for quality control and monitoring in the emulsion-type sausage industry. More importantly, improved nitrite prediction could pave the way for enhanced precision and control in sausage production, ultimately contributing to improved food safety and sustainability.

Is It Possible to Measure the Quality of Sugarcane in Real-Time during Harvesting Using Onboard NIR Spectroscopy?

In-field quality prediction in agricultural products is mainly based on near-infrared spectroscopy (NIR). However, initiatives applied to sugarcane quality are only observed under laboratory-controlled conditions. This study proposed a framework for NIR spectroscopy sensing to measure sugarcane quality during a real harvest operation. A platform was built to support the system composed of the NIR sensor and external lighting on the elevator of a sugarcane harvester. Real-time data were acquired in commercial fields. Georeferenced samples were collected for calibration, validation, and adjustment of the multivariate models by partial least squares (PLS) regression. In addition, subsamples of defibrated cane were NIR-acquired for the development of calibration transfer models by piecewise direct standardization (PDS). The method allowed the adjustment of the spectra collected in real time to predict the quality properties of soluble solids content (Brix), apparent sucrose in juice (Pol), fiber, cane Pol, and total recoverable sugar (TRS). The results of the relative mean square error of prediction (RRMSEP) were from 1.80 to 2.14%, and the ratio of interquartile performance (RPIQ) was from 1.79 to 2.46. The PLS-PDS models were applied to data acquired in real-time, allowing estimation of quality properties and identification of the existence of spatial variability in quality. The results showed that it is possible to monitor the spatial variability of quality properties in sugarcane in the field. Future studies with a broader range of quality attribute values and the evaluation of different configurations for sensing devices, calibration methods, and data processing are needed. The findings of this research will enable a valuable spatial information layer for the sugarcane industry, whether for agronomic decision-making, industrial operational planning, or financial management between sugar mills and suppliers.

Improved quantitative analysis of tenuifolin using hydrolytic continuous-flow system to build prediction models for its content based on near-infrared spectroscopy.

This study used two types of analyses and statistical calculations on powdered samples of Polygala root (PR) and Senega root (SR): (1) determination of saponin content by an independently developed quantitative analysis of tenuifolin content using a flow reactor, and (2) near-infrared spectroscopy (NIR) using crude drug powders as direct samples for metabolic profiling. Furthermore, a prediction model for tenuifolin content was developed and validated using multivariate analysis based on the results of (1) and (2). The goal of this study was to develop a rapid analytical method utilizing the saponin content and explore the possibility of quality control through a wide-area survey of crude drugs using NIR spectroscopy. Consequently, various parameters and appropriate wavelengths were examined in the regression analysis, and a model with a reasonable contribution rate and prediction accuracy was successfully developed. In this case, the wavenumber contributing to the model was consistent with that of tenuifolin, confirming that this model was based on saponin content. In this series of analyses, we have succeeded in developing a model that can quickly estimate saponin content without post-processing and have demonstrated a brief way to perform quality control of crude drugs in the clinical field and on the market.

Near-Infrared Spectroscopy with Supervised Machine Learning as a Screening Tool for Neutropenia.

The use of non-invasive tools in conjunction with artificial intelligence (AI) to detect diseases has the potential to revolutionize healthcare. Near-infrared spectroscopy (NIR) is a technology that can be used to analyze biological samples in a non-invasive manner. This study evaluated the use of NIR spectroscopy in the fingertip to detect neutropenia in solid-tumor oncologic patients. A total of 75 patients were enrolled in the study. Fingertip NIR spectra and complete blood counts were collected from each patient. The NIR spectra were pre-processed using Savitzky-Golay smoothing and outlier detection. The pre-processed data were split into training/validation and test sets using the Kennard-Stone method. A toolbox of supervised machine learning classification algorithms was applied to the training/validation set using a stratified 5-fold cross-validation regimen. The algorithms included linear discriminant analysis (LDA), logistic regression (LR), random forest (RF), multilayer perceptron (MLP), and support vector machines (SVMs). The SVM model performed best in the validation step, with 85% sensitivity, 89% negative predictive value (NPV), and 64% accuracy. The SVM model showed 67% sensitivity, 82% NPV, and 57% accuracy on the test set. These results suggest that NIR spectroscopy in the fingertip, combined with machine learning methods, can be used to detect neutropenia in solid-tumor oncology patients in a non-invasive and timely manner. This approach could help reduce exposure to invasive tests and prevent neutropenic patients from inadvertently undergoing chemotherapy.

Classification of Peruvian Flours via NIR Spectroscopy Combined with Chemometrics

Nowadays, nutritional foods have a great impact on healthy diets. In particular, maca, oatmeal, broad bean, soybean, and algarrobo are widely used in different ways in the daily diets of many people due to their nutritional components. However, many of these foods share certain physical similarities with others of lower quality, making it difficult to identify them with certainty. Few studies have been conducted to find any differences using practical techniques with minimal preparation and in short durations. In this work, Principal Component Analysis (PCA) and Near Infrared Spectroscopy (NIR) were used to classify and distinguish samples based on their chemical properties. The spectral data were pretreated to further highlight the differences among the samples determined via PCA. The results indicate that the raw spectral data of all the samples had similar patterns, and their respective PCA analysis results could not be used to differentiate them. However, pretreated data differentiated the foods in separate clusters according to score plots. The main difference was a C-O band that corresponded to a vibration mode at 4644 cm−1 associated with protein content. PCA combined with spectral analysis can be used to differentiate and classify foods using small samples through the chemical properties on their surfaces. This study contributes new knowledge toward the more precise identification of foods, even if they are combined.

Cerebrovascular reactivity to carbon dioxide tension in newborns: data from combined time-resolved near-infrared spectroscopy and diffuse correlation spectroscopy.

Critically ill newborns are at risk of brain damage from cerebrovascular disturbances. A cerebral hemodynamic monitoring system would have the potential role to guide targeted intervention. To obtain, in a population of newborn infants, simultaneous near-infrared spectroscopy (NIRS)-based estimates of cerebral tissue oxygen saturation () and blood flow during variations of carbon dioxide tension () levels within physiologic values up to moderate permissive hypercapnia, and to examine if the derived estimate of metabolic rate of oxygen would stay constant, during the same variations. We enrolled clinically stable mechanically ventilated newborns at postnatal age without brain abnormalities at ultrasound. and blood flow index were measured using a non-invasive device (BabyLux), which combine time-resolved NIRS and diffuse-correlation spectroscopy. The effect of changes in transcutaneous on , cerebral blood flow (CBF), and cerebral metabolic rate of oxygen index () were estimated. Ten babies were enrolled and three were excluded. Median GA at enrollment was 39 weeks and median weight 2720 g. increased 0.58% (95% CI 0.55; 0.61, ), CBF 2% (1.9; 2.3, ), and 0.3% (0.05; 0.46, ) per mmHg increase in . BabyLux device detected -induced changes in cerebral and CBF, as expected. The small statistically significant positive relationship between and variation is not considered clinically relevant and we are inclined to consider it as an artifact.

Portable NIR spectroscopy: the route to green analytical chemistry.

There is a growing interest for cost-effective and nondestructive analytical techniques in both research and application fields. The growing approach by near-infrared spectroscopy (NIRs) pushes to develop handheld devices devoted to be easily applied for in situ determinations. Consequently, portable NIR spectrometers actually result definitively recognized as powerful instruments, able to perform nondestructive, online, or in situ analyses, and useful tools characterized by increasingly smaller size, lower cost, higher robustness, easy-to-use by operator, portable and with ergonomic profile. Chemometrics play a fundamental role to obtain useful and meaningful results from NIR spectra. In this review, portable NIRs applications, published in the period 2019-2022, have been selected to indicate starting references. These publications have been chosen among the many examples of the most recent applications to demonstrate the potential of this analytical approach which, not having the need for extraction processes or any other pre-treatment of the sample under examination, can be considered the "true green analytical chemistry" which allows the analysis where the sample to be characterized is located. In the case of industrial processes or plant or animal samples, it is even possible to follow the variation or evolution of fundamental parameters over time. Publications of specific applications in this field continuously appear in the literature, often in unfamiliar journal or in dedicated special issues. This review aims to give starting references, sometimes not easy to be found.

Rapid detection of maize seed germination rate based on Gaussian process regression with selection kernel function

Near-infrared spectroscopy (NIR) detection technology enables nondestructive analysis of samples. To achieve fast detection of maize seed germination rates using NIR spectroscopy, 245 samples of maize seeds from three different varieties were purchased from the market and subjected to artificial aging to obtain maize seeds with varying levels of vigor. NIR spectroscopy was collected for each sample, and germination experiments were performed. The Monte Carlo cross-validation method was employed to eliminate 12 outliers, and the baseline correction method along with multiplicative scatter correction (MSC) was applied to the NIR. Partial least squares (PLS), support vector machine (SVM) models, and Gaussian process regression (GPR) models based on different kernel functions were established for detecting maize germination rates. Through comparison of the results of different models, it was found that the GPR model based on the Matern32 kernel function performed best. The R² values for the training and testing sets of this model were 0.9727 and 0.9599, respectively. Experimental results showed that the Matern32-GPR model exhibited greater stability and superior capability in detecting maize seed germination rates. This method can provide a reference for other applications based on NIR detection.

Determination of water-soluble carbohydrates by near-infrared spectroscopy for canola, maize, and sorghum stem fractions

Near infrared spectroscopy (NIRS) was evaluated as a rapid and non-destructive method for determining the concentration of water-soluble carbohydrates (WSC) in stem fractions for winter canola (Brassica napus L.), maize (Zea mays L.) and sorghum (Sorghum bicolor L. Moench) crops. For each crop at different growth stages, stem WSC concentration was determined using NIRS, and benchmarked against the anthrone reagent method, chemical lab analysis. Partial least squares regression was implemented to associate the WSC predicted via NIRS relative to those obtained by laboratory analysis. Spectral regions between 1100 and 1480 nm were critical for WSC determination. The predictive models resulted in coefficient of determinations of 0.93, 0.94, and 0.95, and a Root Mean Square Error of prediction of 10, 20, and 17 for winter-canola, maize and sorghum crops, respectively. The NIRS spectroscopy is a reliable method for WSC determination in stem tissues on these major field crops.

Application of Near Infrared Spectroscopy for the Rapid Assessment of Nutritional Quality of Different Strawberry Cultivars.

Strawberry is the most cultivated berry fruit globally and it is really appreciated by consumers because of its characteristics, mainly bioactive compounds with antioxidant properties. During the breeding process, it is important to assess the quality characteristics of the fruits for a better selection of the material, but the conventional approaches involve long and destructive lab techniques. Near infrared spectroscopy (NIR) could be considered a valid alternative for speeding up the breeding process and is not destructive. In this study, a total of 216 strawberry fruits belonging to four different cultivars have been collected and analyzed with conventional lab analysis and NIR spectroscopy. In detail, soluble solid content, acidity, vitamin C, anthocyanin, and phenolic acid have been determined. Partial least squares discriminant analysis (PLS-DA) models have been developed to classify strawberry fruits belonging to the four genotypes according to their quality and nutritional properties. NIR spectroscopy could be considered a valid non-destructive phenotyping method for monitoring the nutritional parameters of the fruit and ensuring the fruit quality, speeding up the breeding program.

Rapid and Green Classification Method of Bacteria Using Machine Learning and NIR Spectroscopy.

Green Chemistry is a vital and crucial instrument in achieving pollution control, and it plays an important role in helping society reach the Sustainable Development Goals (SDGs). NIR (near-infrared spectroscopy) has been utilized as an alternate technique for molecular identification, making the process faster and less expensive. Near-infrared diffuse reflectance spectroscopy and Machine Learning (ML) algorithms were utilized in this study to construct identification and classification models of bacteria such as Escherichia coli, Salmonella enteritidis, Enterococcus faecalis and Listeria monocytogenes. Furthermore, divide these bacteria into Gram-negative and Gram-positive groups. The green and quick approach was created by combining NIR spectroscopy with a diffuse reflectance accessory. Using infrared spectral data and ML techniques such as principal component analysis (PCA), hierarchical cluster analysis (HCA) and K-Nearest Neighbor (KNN), It was feasible to accomplish the identification and classification of four bacteria and classify these bacteria into two groups: Gram-positive and Gram-negative, with 100% accuracy. We may conclude that our study has a high potential for bacterial identification and classification, as well as being consistent with global policies of sustainable development and green analytical chemistry.

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.

Direct NIR spectral determination of genetic improvement, light availability, and their interaction effects on chemically selected yerba-mate leaves

Monitoring plant chemical modifications provides useful information about plant genetic variability owing to environmental stresses. This allows determining if the pattern of chemical compounds of interest remains constant. Classically, chemical compounds are monitored through classical analytical procedures that are costly and environmentally harmful. Decaffeinated and high caffeine yerba-mate trees were selected based on their chemical profiles. Their behavior was evaluated at five different shading levels (0%, 40%, 51%, 76%, and 82%), imitating the environmental characteristics of yerba-mate in agroforestry systems. Near-infrared spectroscopy (NIR) fingerprints were obtained for yerba-mate leaves, and ANOVA-simultaneous component analysis (ASCA) models were developed for chemical monitoring. The preprocessing of NIR spectra was assessed with a three-level full factorial design. The best preprocessing set of conditions found was standard normal variate followed by second derivatives, computed according to the Savitzky–Golay method (11-points window and second-degree interpolating polynomial). The ASCA models satisfactorily showed significant differences in chemical behavior for genetic improvement and its interaction with light availibility effects. The genetic improvement effect had a variance of 44.30% over NIR fingerprints, indicating that the chemical differences are maintained at any light level. The interaction effect with 1.84% of the total variance illustrated that the behaviors of decaffeinated and high-caffeinated yerba mate were different when subjected to the same light level. The main absorptions indicated by the ASCA model were the O-H stretching and bending combination band at 1654 nm, and the –CH stretching aromatic 1st overtone at 1675 nm. The ASCA method offered a powerful tool for monitoring chemically selected plants combined with NIR spectroscopy. The method is cheap, fast, environmentally friendly, and can be used for field research.

Non-targeted authentication of black pepper using a local web platform: Development, validation and post-analytical challenges of a combined NIR spectroscopy and LASSO method

Black pepper is a valuable commodity susceptible to economically motivated adulterations. This contribution describes the proper development, standardization, validation and late-stage validation of a non-targeted method for authentication of black pepper by near-infrared spectroscopy (NIR) coupled to a least absolute shrinkage and selection operator (LASSO). Note that this is the first successful attempt to apply the LASSO method to infrared spectroscopy data. We analysed 150 diverse samples of black pepper supplied by a well-recognized spice trader. A first batch of samples (n = 116) were split into training and test sets and then normalized by multiplicative scatter correction (MSC). While the test set was withheld for further testing of the model, the training set was submitted to the LASSO method with the aim of retrieving the discriminant spectral features and classifying the samples as authentic, exogenously-adulterated or endogenously-adulterated. The model was tested on the test set, achieving an overall accuracy of 94% with very high sensitivity and specificity rates. Furthermore, an R-based local web platform was created for immediate prediction of the samples from raw data. The NIR user simply uploads the raw data to the local web application and MSC normalization on the stored training set and interrogation of the LASSO classifier are performed. The online application facilitates the model testing and enables clear visualization of the outcomes. After the creation of the local web platform, the classifier was then validated with a new batch of independent samples, resulting in the correct prediction of 33/34 samples. The robustness of this non-targeted method and its stability over time was further established, and a late-stage validation of the classifier was successfully performed by an inexperienced user. Finally, the LASSO classifier was successfully challenged with a sample from an international inter-laboratory proficiency test. • A non-targeted method for authentication of black pepper by NIR & LASSO. • A web application was created for direct classification of NIR raw data. • Validation and late-stage challenges of the method were carried out. • Successful participation to an international proficiency test.

Combination of NIR spectroscopy and algorithms for rapid differentiation between one-year and two-year stored rice

Storage is necessary for rice to ensure the year-round consumption of rice. With the increase in storage time, the taste quality and commercial value of rice gradually decrease. The accurate determination of the freshness of rice is critical to the rice trade. However, it is difficult to distinguish aging rice from fresh rice, so a quick and simple method is needed to identify the freshness of the rice. In this study, a combination of near-infrared spectroscopy (NIR) and various algorithms, such as partial least squares discriminant analysis (PLS-DA), support vector machines (SVM), and classification and regression trees (CART), were used to differentiate the freshness of rice. PLS-DA and SVM demonstrated excellent classification ability in identifying the freshness of rice, with sensitivity and specificity of 1. The original spectra were used with 100% accuracy in the test set to determine the freshness of the rice. As a result, PLS-DA and SVM can be used to determine the freshness of the rice.

Prediction of Soluble Solids Content by Means of NIR Spectroscopy and Relation with Botrytis cinerea Tolerance in Strawberry Cultivars

Strawberry fruits are particularly appreciated by consumers for their sweet taste related to their soluble solids content (SSC). However, strawberries are characterized by a short shelf-life and high susceptibility to tissue infection, mainly by Botrytis cinerea. The SSC determination of strawberry fruit through traditional destructive techniques has some limitations related to the applicability, timing, and number of samples. The aims of this study are (i) to verify if any relation between SSC and B. cinerea susceptibility in the fruits of five strawberry cultivars occurs and (ii) to determine the SSC of strawberry fruits through near infrared spectroscopy (NIR). Principal component analysis was used to search for spectral differences among the strawberry genotypes. The partial least squares regression technique was computed in order to predict the SSC of the fruits collected during two harvesting seasons. Moreover, variable selection methods were tested in order to improve the models and get better predictions. The results demonstrated that there was a high correlation between SSC and B. cinerea susceptibility (R2 up to 0.87). The SSC was predicted with a standard error of 0.84 °Brix and R2p 0.75 (for the best model), which indicated the possibility to use the models for screening applications. NIR spectroscopy represents an important non-destructive alternative and finds remarkable applications in the agro-food market.

Sex classification of silkworm pupae from different varieties by near infrared spectroscopy combined with chemometrics

The accuracy of sex classification of silkworm pupae is directly related to the quality of crossbred silkworm pupae and silkworm silk. Manual selection as the traditional method for sex identification of silkworm pupae is labor-intensive and time-consuming. In addition, there are hundreds of silkworm pupae varieties. Currently, the widely used near infrared spectroscopy（NIR) technology showed unsatisfactory accuracy in distinguishing the sex of different species of silkworm pupae. Thus, we developed a method combining NIR and contrastive principal component analysis(cPCA) to achieve high precision sex classification of different silkworm pupae by introducing the background data set to eliminate the influence of silkworm pupae varieties. The experimental results showed that the cPCA analysis showed better performance (98.34%) compared with the accuracy of PCA analysis (71.36%). After characteristic wavenumber selection by iPLS, the accuracy could be improved to 99.58% by using the cPCA algorithm. The combination of NIR spectroscopy with cPCA provides a method for developing portable devices to quickly identify the sex of silkworm pupae.