Near-infrared Spectroscopy Prediction Research Articles

Combination of near-infrared spectroscopy with Wasserstein generative adversarial networks for rapidly detecting raw material quality for formula products.

Near-infrared spectroscopy (NIRS) has emerged as a key technique for rapid quality detection owing to its fast, non-destructive, and eco-friendly characteristics. However, its practical implementation within the formulation industry is challenging owing to insufficient data, which renders model fitting difficult. The complexity of acquiring spectra and spectral reference values results in limited spectral data, aggravating the problem of low generalization, which diminishes model performance. To address this problem, we introduce what we believe to be a novel approach combining NIRS with Wasserstein generative adversarial networks (WGANs). Specifically, spectral data are collected from representative samples of raw material provided by a formula enterprise. Then, the WGAN augments the database by generating synthetic data resembling the raw spectral data. Finally, we establish various prediction models using the PLSR, SVR, LightGBM, and XGBoost algorithms. Experimental results show the NIRS-WGAN method significantly improves the performance of prediction models, with R2 and RMSE of 0.949 and 1.415 for the chemical components of sugar, respectively, and 0.922 and 0.243 for nicotine. The proposed framework effectively enhances the predictive capabilities of various models, addressing the issue caused by limited training data in NIRS prediction tasks.

Application of quantitative non-destructive determination of protein in wheat based on pretreatment combined with parallel convolutional neural network

With the increasing demand for wheat, the detection of wheat quality has become imperative. Protein content is an important indicator for wheat quality. Near infrared spectroscopy (NIRS) quantitative non-destructive testing technology has gained widespread application in agricultural field with the development of science and chemometrics technology. In this study, NIRS system was employed to measure the spectra of wheat, and the original spectra were pretreated using Savitzky-Golay smoothing (SG) pretreatment method. Subsequently, the NIRS prediction model of protein in wheat that using SG combined with parallel convolutional neural network (PaBATunNet) was established. PaBATunNet was composed of a one-dimensional convolutional layer, a parallel convolution module (Module), a flattening layer, four fully connected layers and a parameter regulator (PR). Module was made up of five submodules and a Concatenate function. The multidimensional features of the spectra were extracted by five submodules and spliced by Concatenate function. SG pretreatment combined with PaBATunNet (SG-PaBATunNet) was compared with commonly modeling methods, such as SG-partial least squares (SG-PLS), SG-principal component regression (SG-PCR), SG-support vector machine (SG-SVM) and SG-back propagation neural network (SG-BP). The results demonstrated that the modeling accuracy and prediction accuracy of SG-PaBATunNet were improved by 26.7%, 23.9%, 45.6%, 44.2%, and 38.4%, 39.6%, 60.1%, 58.0%, when compared with SG-PLS, SG-PCR, SG-SVM and SG-BP. The problems of low prediction accuracy and poor generalization ability with commonly modeling methods were effectively addressed by SG-PaBATunNet. This study provides an essential theoretical foundation for developing a fast, nondestructive and high-precision NIRS quantitative analysis model of protein in wheat.

On-Line Monitoring of Enzymatic Degumming of Soybean Oil Using Near-Infrared Spectroscopy.

Degumming is an oil refinement process in which the naturally occurring phospholipids in crude vegetable oils are removed. Enzymatic degumming results in higher oil yield and more cost-efficient processing compared to traditional degumming processes using only water or acid. Phospholipase C hydrolyses phospholipids into diglycerides and phosphate groups during degumming. The diglyceride content can therefore be considered a good indicator of the state of the enzymatic reaction. This study investigates the use of near-infrared (NIR) spectroscopy and chemometrics to monitor the degumming process by quantifying diglycerides in soybean oil in both off-line and on-line settings. Fifteen enzymatic degumming lab scale batches originating from a definitive screening design (with varying water, acid, and enzyme dosages) were investigated with the aim to develop a NIR spectroscopy prediction method. By applying tailored preprocessing and variable selection methods, the diglyceride content can be predicted with a root mean square error of prediction of 0.06% (w/w) for the off-line set-up and 0.07% (w/w) for the on-line set-up. The results show that the diglyceride content is a good indicator of the enzyme performance and that NIR spectroscopy is a suitable analytical technique for robust real-time diglyceride quantification.

Prediction of Anthocyanidins Content in Purple Chinese Cabbage Based on Visible/Near Infrared Spectroscopy.

Purple Chinese cabbage (PCC) has become a new breeding trend due to its attractive color and high nutritional quality since it contains abundant anthocyanidins. With the aim of rapid evaluation of PCC anthocyanidins contents and screening of breeding materials, a fast quantitative detection method for anthocyanidins in PCC was established using Near Infrared Spectroscopy (NIR). The PCC samples were scanned by NIR, and the spectral data combined with the chemometric results of anthocyanidins contents obtained by high-performance liquid chromatography were processed to establish the prediction models. The content of cyanidin varied from 93.5 mg/kg to 12,802.4 mg/kg in PCC, while the other anthocyanidins were much lower. The developed NIR prediction models on the basis of partial least square regression with the preprocessing of no-scattering mode and the first-order derivative showed the best prediction performance: for cyanidin, the external correlation coefficient (RSQ) and standard error of cross-validation (SECV) of the calibration set were 0.965 and 693.004, respectively; for total anthocyanidins, the RSQ and SECV of the calibration set were 0.966 and 685.994, respectively. The established models were effective, and this NIR method, with the advantages of timesaving and convenience, could be applied in purple vegetable breeding practice.

Precise control of water stress in the field reveals different response thresholds for forage yield and digestibility of maize hybrids.

With dwindling global freshwater supplies and increasing water stress, agriculture is coming under increasing pressure to reduce water use. Plant breeding requires high analytical capabilities. For this reason, near-infrared spectroscopy (NIRS) has been used to develop prediction equations for whole-plant samples, particularly for predicting dry matter digestibility, which has a major impact on the energy value of forage maize hybrids and is required for inclusion in the official French catalogue. Although the historical NIRS equations have long been used routinely in seed company breeding programmes, they do not predict all variables with the same accuracy. In addition, little is known about how accurate their predictions are under different water stress-environments. Here, we examined the effects of water stress and stress intensity on agronomic, biochemical, and NIRS predictive values in a set of 13 modern S0-S1 forage maize hybrids under four different environmental conditions resulting from the combination of a northern and southern location and two monitored water stress levels in the south. First, we compared the reliability of NIRS predictions for basic forage quality traits obtained using the historical NIRS predictive equations and the new equations we recently developed. We found that NIRS predicted values were affected to varying degrees by environmental conditions. We also showed that forage yield gradually decreased as a function of water stress, whereas both dry matter and cell wall digestibilities increased regardless of the intensity of water stress, with variability among the tested varieties decreasing under the most stressed conditions. By combining forage yield and dry matter digestibility, we were able to quantify digestible yield and identify varieties with different strategies for coping with water stress, raising the exciting possibility that important potential selection targets still exist. Finally, from a farmer's perspective, we were able to show that late silage harvest has no effect on dry matter digestibility and that moderate water stress does not necessarily result in a loss of digestible yield.

Rapid detection of common wheat flour addition to durum wheat flour and pasta using spectroscopic methods and chemometrics

Attenuated total reflectance Fourier-transform infrared spectroscopy (ATR-FTIRS), near-infrared spectroscopy (NIRS), and synchronous fluorescence spectroscopy (SFS) combined with chemometrics were used to determine common wheat (Triticum aestivum) flour in durum wheat (Triticum durum) flour and common wheat farina addition to durum wheat pasta. The main objective of this study was to determine the most suitable spectroscopic method(s) to predict adulteration level in flour and pasta. It was found that among the spectroscopic methods tested, ATR-FTIRS had the lowest limit of detection (LOD) values for blended flour (0.68%) and pasta samples (0.49%). The coefficients of determination (R2) values of the prediction curves for ATR-FTIRS were found as 0.900 and 0.903 for flour and pasta samples, respectively. Similarly, R2 for the prediction of NIRS was 0.867 and 0.895 for flour and pasta samples, respectively. Since, NIR spectroscopy is widely used in food industry for quality control, it might have a good potential to detect the adulteration of common wheat flour in durum wheat flour and farina in the durum wheat pasta. SFS demonstrated a weaker predictive ability for low levels of adulteration with its higher root mean square error of prediction (10.750). The study demonstrated that ATR-FTIRS had better specificity and accuracy compared to other spectroscopic methods for the detection of adulteration.

A near‐infrared spectroscopy method for the detection of texture profile analysis of Litopeneo vannamei based on partial least squares regression

AbstractAmong aquatic products freshness testing methods, spectral detection method is the most commonly used rapid nondestructive testing method, and the quality structure index can be used as the judgment index of aquatic products freshness, so studying the near‐infrared (NIR) spectroscopy detection method of aquatic products based on texture index is helpful to improve the detection accuracy of freshness of products. In order to find a method for NIR spectroscopy based on texture indicators, we took shrimp as the research target, collected total 216 NIR spectral data of 680–2600 nm and texture index of sample for 8 consecutive days, and established a quantitative prediction model. In this study, we compared the effects of six spectral pretreatments on partial least squares regression (PLSR) models, including multiple scattering correction, standard normal variate transformation, mean centering, Savitzky–Golay (S‐G) smoothing, moving average method, and normalization. The results show that the PLSR model based on S‐G smoothing has the best model accuracy in prediction after the S‐G smoothing algorithm has preprocessed the spectral data. In addition, principal component analysis and the Mahalanobis distance algorithm were used in the study to remove abnormal samples to improve the accuracy of the model. Finally, the quantitative prediction model for qualitative construct indicators constructed based on the S‐G smoothing algorithm and the PLSR algorithm had better prediction results than the other models algorithm we considered, with a correlation coefficient (R) of 0.868 for the cohesive validation set and 0.781 for the elastic validation set of qualitative construct indicators. Finally, based on the above study, this study developed a prediction graphical user interface system for qualitative metrics of texture.Practical ApplicationsMost of the existing aquatic meat freshness detection methods are offline detection based on physical and chemical tests, and some use spectral technology to detect the chemical parameters of aquatic meat freshness. This project innovatively applies NIR spectroscopy technology to aquatic products. The physical parameters of meat freshness can be quickly detected online. By preprocessing the data samples and selecting the response characteristic wavelengths, a NIR spectroscopy prediction model of texture indicators with self‐learning ability and high precision is constructed, and the texture is established based on this. Indicator NIR spectroscopy prediction system. This whole process provides new research methods and analytical tools for rapid nondestructive testing of aquatic product freshness.

Algorithm of Stability-Analysis-Based Feature Selection for NIR Calibration Transfer.

For conventional near-infrared spectroscopy (NIR) technology, even within the same sample, the NIR spectral signal can vary significantly with variation of spectrometers and the spectral collection environment. In order to improve the applicability and application of NIR prediction models, effective calibration transfer is essential. In this study, a stability-analysis-based feature selection algorithm (SAFS) for NIR calibration transfer is proposed, which is used to extract effective spectral band information with high stability between the master and slave instruments during the calibration transfer process. The stability of the spectrum bands shared between the master and slave instruments is used as the evaluation index, and the genetic algorithm was used to select suitable thresholds to filter out the spectral feature information suitable for calibration transfer. The proposed SAFS algorithm was applied to two near-infrared datasets of corn oil content and larch wood density. Simultaneously, its calibration transfer performances were compared with two classical feature selection methods. The effects of different preprocessing algorithms and calibration transfer algorithms were also assessed. The model with the feature variables selected by the SAFS obtained the best prediction. The SAFS algorithm can simplify the spectral data to be transferred and improve the transfer efficiency, and the universality of the SAFS allows it to be used to optimize calibration transfer in various situations. By combining different preprocessing and classic feature selection methods with this, the sensitivity of the correlation between spectral data and component information are improved significantly, as well as the effect of calibration transfer, which will be deeply developed.

Near-Infrared Spectroscopy for Prediction of Potentially Toxic Elements in Soil and Sediments from a Semiarid and Coastal Humid Tropical Transitional River Basin

Near-Infrared Spectroscopy for Prediction of Potentially Toxic Elements in Soil and Sediments from a Semiarid and Coastal Humid Tropical Transitional River Basin

Estimation of apparent dietary nutrient digestibility in Yellowtail Kingfish Seriola lalandi by Near-Infrared Spectroscopy (NIRS)

The apparent digestibility (AD) of dietary nutrients is an essential criterion used to improve nutrition in aquaculture. We assessed whether Near-Infrared Spectroscopy (NIRS) can assist in streamlining the process by predicting directly diet AD or the composition of diets and/or faecal nutrients and digestibility marker (yttrium oxide) concentrations collected in Yellowtail Kingfish (YTK) in vivo experiments. NIRS models were developed using up to 121 diets and 185 faecal samples, and assessed as part of a case-study investigating the accuracy of AD results for 7 validation diets using a range of prediction scenarios (S2-S5) against the benchmark analytical method (S1; all samples analysed). A large range in AD of diets was noted and results from the case-study suggested diet nutrient digestibility was significantly affected by the type of alternative ingredients used (AD dry matter ranged 45–73%; AD protein 75–85%; AD lipid 66–92%; AD energy 65–87%; AD amino acids 78–86%; AD methionine 81–90%; AD lysine 77–91%; AD taurine 70–79%). NIRS models had cross-validation R2 values ranging 0.63–0.89 and residual predictive deviation (RPD) values ranging 0.7–3.3 for diet nutrients; R2 values ranging 0.63–0.94 and RPD values ranging 1.8–4.0 for faeces. Successful NIRS models were developed for diet dry matter, ash, crude protein, total lipid, gross energy, and total amino acids; successful faeces models also included methionine and lysine but not dry matter and ash; taurine and yttrium oxide models were deemed unsuccessful for both diets and faeces. Dry matter AD based on either yttrium concentration predictions (S4) or direct prediction of diet AD (S5) were relatively inaccurate (±20% error in dry matter AD). Nonetheless S4 was relatively successful to predict lipid AD and S5 the protein AD of the seven diets with an average error of ±6.7% AD and ± 5% AD respectively compared to the benchmark. When the digestibility marker concentrations were analysed, the S2 (diet composition predicted) and S3 (diet and faeces composition predicted) scenarios could predict nutrient AD with a degree of error of ±1–3%, with the exception of lipid AD (±6–8% error), and taurine which was modelled poorly in this study. NIRS prediction of diet and faecal composition, combined with digestibility marker analysis, appears to be a cost-effective and reasonably accurate rapid method to assess diet AD in fish, reducing sample volume requirements, welfare impact and opening a range of new opportunities for fish nutrition and the development of more efficient phenotypes.

Prediction performance optimization of different resolution and spectral band ranges for characterizing coco-peat substrate available nitrogen

To realize the high-precision measurement of available nitrogen in coco-peat substrate, the influences of near-infrared (NIR) spectral data of different spectral resolution and band ranges on the model detection performances were studied. A flame NIR spectrometer and an ANTARIS FT-NIR spectrometer were used to acquire spectral data of short-wave near-infrared (SW-NIR) band (1000–1660 nm), long-wave near-infrared (LW-NIR) band (1660–2500 nm), and NIR band (1000–2500 nm) of coco-peat substrate samples respectively. Then, the models were established based on the original data and various pre-processing data. Furthermore, the effects of spectral resolution and detection band ranges on available nitrogen NIR spectroscopy detection of coco-peat substrate were studied by comparing the modeling results. The spectral resolution and detection band ranges had significant influences on the NIR spectroscopy detection of coco-peat substrate available nitrogen. In comparison, the model with high-resolution LW-NIR band spectral data yielded the best prediction performance. Specifically, the correlation coefficients and root mean squared errors in calibration set and prediction set were 0.999, 0.996, 3.197 mg/100g, and 5.373 mg/100g, respectively. The ratio of prediction to deviation was 10.895. High-resolution spectral data was useful to improve the model performance of the sample available nitrogen NIR spectroscopy prediction, and the LW-NIR band range was the optimal detection band for coco-peat substrate available nitrogen NIR spectroscopy detection. Therefore, the high-resolution LW-NIR band spectrometer was useful to the development of high-precision NIR spectroscopy detection equipment for available nitrogen in coco-peat substrate.

The accuracy of NIRS in predicting chemical composition and fibre digestibility of hay-based total mixed rations

The aim of this study was to develop near-infrared spectroscopy (NIRS) prediction models for the estimation of chemical components and the fibre undegradable fractions (uNDF) of hay-based total mixed rations (TMR). A total of 205 TMR samples were used for the study. All the chemical components were measured using standard AOAC reference methods and expressed as percentages of dry matter (DM). Prediction models were developed using both cross- and independent validation and different mathematical treatments applied on spectral data. The best spectral treatment was chosen based on the method which simultaneously achieved the lowest root mean square error and the highest explained variance in cross-validation. The coefficient of determination in external validation (R2 P) was the greatest for starch prediction model (R2 P = 0.84), followed by acid detergent fibre (ADF; R2 P = 0.79), and amylase-treated ash-corrected NDF with addition of sodium sulphite (aNDFom) and crude protein prediction models (CP; R2 P = 0.73). The concordance correlation coefficient (CCC) in validation ranged from 0.66 (ash prediction model) to 0.92 (starch prediction model), indicating substantial to accurate models’ predictive ability. This study indicated that NIRS can be a screening method for the prediction of CP, Starch, aNDFom, ADF, acid detergent lignin (ADL), uNDF and Ash. The use of TMR utilised in various herds provided high variability for the NIRS calibration dataset, implying that the developed NIRS pre-diction models could be applicable to TMR collected from herds located in the Parmigiano Reggiano cheese production area. Highlights NIRS can be successfully employed to determine quickly and at cost-effective different compositional and digestibility traits in hay-based TMR. TMR analysis predicted by NIRS can support nutritionists in the formulation of diets containing a proper nutrient profile to sustain physiological, metabolic, and immunological processes. The use of NIR technology for TMR analysis can allow frequent monitoring of rations and increasingly timely corrections, maximising cows’ diet utilisation and conversion of the ingested feed.

Industrial-scale continuous on-line analysis of liquid hog manure by NIRS

The objective of the work was to use near-infrared spectroscopy (NIRS) to identify the pattern of distribution of the manure nutrients onto a field. The paper describes what is believed to be the first application of NIRS to continuous analysis of liquid manure on an industrial scale. The on-line analysis is accomplished without the need for a sample cell to scan the sample, or the need for sampling during manure application to the land, because the entire amount of the material is analyzed. Diode array instruments were used for continuous scanning of the manure as it was being applied. Because of differences among the instruments in spectral data, four instruments were set up in line for the development of the industrial-level calibration. Spectral data were recorded as direct reflectance over the range of 11,100 cm−1 to 5,900 cm−1. Calibrations for total solids, nitrogen, phosphorus, and potassium were developed using the Unscrambler, with a data pre-treatment of first derivative. No outliers were removed. Values of the RPD (ratio of SD of reference data to SEP) of between 3.1 and 4 were attained for all constituents. Interpolation of blind duplicates revealed that reproducibility of NIRS predictions for all constituents was significantly superior to that of laboratory analysis by certified laboratories. This was attributed to the enormous differences in the volume of manure scanned by the NIRS instruments, relative to the volumes used in laboratory analysis. Continuous scanning and analysis of liquid manures enable mapping of nutrients as they are being applied to the land, which provides valuable information to farmers as to the purchase, and efficiency of use of fertilizers.

Predicting the cell-wall compositions of solid Pinus radiata (radiata pine) wood using NIR and ATR FTIR spectroscopies

Infrared spectroscopy coupled with partial least squares (PLS) regression has been shown to be a rapid alternative to wet chemical analytical methods for determining the cell-wall compositions of wood. Both near infrared (NIR) spectroscopy, and mid-infrared spectroscopy with attenuated total reflectance Fourier transform infrared (ATR FTIR) sampling, coupled with PLS regression, can be used to quickly and accurately predict the lignin contents and monosaccharide compositions of milled wood. However, milling wood can be time consuming and laborious. In this study we demonstrate that PLS-1 models built using NIR and ATR FTIR spectra of milled Pinus radiata wood, with different sized wood particles and different moisture contents, can rapidly and accurately predict the cell-wall compositions of solid wood. A robust assessment of the prediction accuracy was conducted using a separate test set of solid wood samples with both ‘smooth’ and ‘rough’ surface finishes. The lowest standard error (SE) values for most of the compositional predictions were obtained for the ‘rough’ solid wood samples, using PLS-1 models built from NIR spectra of ‘large’ milled wood particles (0.422 mm) with ambient moisture content. The SE achieved for NIR spectroscopy prediction of lignin for the ‘rough’ solid wood was 1.91%, and for the monosaccharides, arabinose (0.37%), xylose (1.25%), galactose (2.00%), mannose (1.54%), and 4-O-methyl glucuronic acid (0.24%). The powerful combination of NIR spectroscopy with PLS regression offers an attractive method for rapid prediction of cell-wall compositions of solid wood samples, thus avoiding milling. In addition, this technique highlights the different levels of these cell-wall components in opposite and compressed regions in solid wood.

Fine root lignin content is well predictable with near-infrared spectroscopy

1. Root lignin is a key driver of root decomposition, which in turn is a fundamental component of the terrestrial carbon cycle and increasingly in the focus of ecologists and global climate change research. However, measuring lignin content is labor-intensive and therefore not well-suited to handle the large sample sizes of most ecological studies. To overcome this bottleneck, we explored the applicability of high-throughput near infrared spectroscopy (NIRS) measurements to predict fine root lignin content. 2. We measured fine root lignin content in 73 plots of a field biodiversity experiment containing a pool of 60 grassland species using the Acetylbromid (AcBr) method. To predict lignin content, we established NIRS calibration and prediction models based on partial least square regression (PLSR) resulting in moderate prediction accuracies (RPD = 1.96, R2 = 0.74, RMSE = 3.79). 3. In a second step, we combined PLSR with spectral variable selection. This considerably improved model performance (RPD = 2.67, R2 = 0.86, RMSE = 2.78) and enabled us to identify chemically meaningful wavelength regions for lignin prediction. 4. We identified 38 case studies in a literature survey and quantified median model performance parameters from these studies as a benchmark for our results. Our results show that the combination Acetylbromid extracted lignin and NIR spectroscopy is well suited for the rapid analysis of root lignin contents in herbaceous plant species even if the amount of sample is limited.

The adaptable use of Brassica NIRS calibration equations to identify pennycress variants to facilitate the rapid domestication of a new winter oilseed crop

Pennycress is being domesticated as a new winter oilseed crop to be grown between corn harvest and soybean planting the following year in the Upper Midwestern United States. The aim of this research was to evaluate seed composition traits in a large pennycress mutant population using near-infrared spectroscopy (NIRS). We tested the hypothesis that Brassica NIRS calibration equations could be used rapidly and cost effectively to evaluate pennycress seeds for protein, oil, glucosinolate, and fatty acid contents. Using calibration equations developed for Brassica, we identified a broad range for each of the traits evaluated in this study. Wet lab analyses were conducted on selected subsets of the mutant families to confirm the predicted variation for each trait using NIRS. Supporting our hypothesis, moderate to strong correlations were observed between the wet lab analyses and NIRS predictions for seven of the eight traits evaluated. For protein, oil and glucosinolate content, correlations between the NIRS predictions and wet lab values were 0.82, 0.92 and 0.78 respectively. For fatty acid content, moderate correlations for oleic acid (0.68), linoleic acid (0.78), linolenic acid (0.60) and erucic acid (0.74) were observed. These findings allowed us to use NIRS to quickly evaluate large heterogenous mutant families and identify pennycress lines harboring desirable traits required for adoption as a cover crop and to create several industrial opportunities.

Visible‐Near‐Infrared Spectroscopy Prediction of Soil Characteristics as Affected by Soil‐Water Content

Core Ideas The soil–water retention characteristics govern the water effect on soil vis‐NIR spectra. Soil clay, silt, sand content, and the water content at pF3 can be predicted with high degree of accuracy using spectroscopy at any field moisture condition. The OC and the derived properties of OC showed poorer prediction when the soil–water content exceeds field capacity (pF<2). Soil physical characteristics are important drivers for soil functions and productivity. Field applications of near‐infrared spectroscopy (NIRS) are already deployed for in situ mapping of soil characteristics and therefore, fast and precise in situ measurements of the basic soil physical characteristics are needed at any given water content. Visible‐near‐infrared spectroscopy (vis–NIRS) is a fast, low‐cost technology for determination of basic soil properties. However, the predictive ability of vis–NIRS may be affected by soil‐water content. This study was conducted to quantify the effects of six different soil‐water contents (full saturation, pF 1, pF 1.5, pF 2.5, pF 3, and air‐dry) on the vis–NIRS predictions of six soil physical properties: clay, silt, sand, water content at pF 3, organic carbon (OC), and the clay/OC ratio. The effect of soil‐water content on the vis–NIR spectra was also assessed. Seventy soil samples were collected from five sites in Denmark and Germany with clay and OC contents ranging from 0.116 to 0.459 and 0.009 to 0.024 kg kg‐1, respectively. The soil rings were saturated and successively drained/dried to obtain different soil–water potentials at which they were measured with vis–NIRS. Partial least squares regression (PLSR) with leave‐one‐out cross‐validation was used for estimating the soil properties using vis–NIR spectra. Results showed that the effects of water on vis–NIR spectra were dependent on the soil–water retention characteristics. Contents of clay, silt, and sand, and the water content at pF 3 were well predicted at the different soil moisture levels. Predictions of OC and the clay/OC ratio were good at air‐dry soil condition, but markedly weaker in wet soils, especially at saturation, at pF 1 and pF 1.5. The results suggest that in situ measurements of spectroscopy are precise when soil‐water content is below field capacity.

Assessing the influence of temperature on NIRS prediction models for the determination of sodium content in dry-cured ham slices

Temperature fluctuations are a key factor in the development of prediction models using near infrared spectroscopy (NIRS). In the present study, this influence has been investigated and a methodology has been proposed to reduce the effect of sample temperature on NIRS model prediction of the sodium content in dry-cured ham slices. Spectra were taken directly from the slices using a remote measurement probe (for non-contact analysis) at three different temperature ranges: −12 °C to −5°C, −5°C to 10 °C and 10 °C to 20 °C. Local and global temperature compensation methods were established. Partial-least squares (PLS) regression was used as a chemometrics tool to perform the calibrations. The results showed that local models were sensitive to changes in temperature, while a global temperature model using sample spectra over the entire temperature range showed good prediction ability, reducing the error caused by temperature fluctuations to acceptable levels for practical applications.

Utilizing functional near-infrared spectroscopy for prediction of cognitive workload in noisy work environments.

.Occupational noise frequently occurs in the work environment in military intelligence, surveillance, and reconnaissance operations. This impacts cognitive performance by acting as a stressor, potentially interfering with the analysts’ decision-making process. We investigated the effects of different noise stimuli on analysts’ performance and workload in anomaly detection by simulating a noisy work environment. We utilized functional near-infrared spectroscopy (fNIRS) to quantify oxy-hemoglobin (HbO) and deoxy-hemoglobin concentration changes in the prefrontal cortex (PFC), as well as behavioral measures, which include eye tracking, reaction time, and accuracy rate. We hypothesized that noisy environments would have a negative effect on the participant in terms of anomaly detection performance due to the increase in workload, which would be reflected by an increase in PFC activity. We found that HbO for some of the channels analyzed were significantly different across noise types (). Our results also indicated that HbO activation for short-intermittent noise stimuli was greater in the PFC compared to long-intermittent noises. These approaches using fNIRS in conjunction with an understanding of the impact on human analysts in anomaly detection could potentially lead to better performance by optimizing work environments.

DETERMINATION OF BRACHIARIA SPP. FORAGE QUALITY BY NEAR-INFRARED SPECTROSCOPY AND PARTIAL LEAST SQUARES REGRESSION

Characterizing the chemical properties of forage is critical for the production of improved pastures and livestock development. Conventional analysis methods are very time- and material-consuming, whereas near-infrared spectroscopy (NIRS) and chemometric analyses allow a fast simultaneous determination of various chemical or physical properties without the use of solvents or large sample amounts. The present research involved the development of models based on NIRS and partial least squares regression (PLS) to estimate the neutral detergent fiber (NDF), acid detergent fiber (ADF), cellulose, and crude protein (CP) contents in Brachiaria spp. forage samples. The models were constructed using spectral data in the range of 800 to 1850 nm. Different preprocessing methods were applied, such as standard normal variate and first-/second-derivative transformations. The obtained calibration models were internally cross-validated, displaying validation errors similar to those obtained for conventional methods. The predictive abilities of the developed models were evaluated for external set samples. NDF, ADF, cellulose, and CP contents were estimated with relative errors of prediction (REPs) of 1.8, 2.6, 4.1, and 8.5%, respectively. NIRS predictions are a useful and profitable tool for fast multi-sample chemical property analysis that is required for the assessment of forage quality. The obtained models are suitable for estimating the key chemical characteristics of forage quality. This research contributes a new approach to determining the quality of Brachiaria spp. forage and provides a new technological tool for the improvement of this crop.