NIR Data Research Articles

Mid-Level Data Fusion Combined with the Fingerprint Region for Classification DON Levels Defect of Fusarium Head Blight Wheat.

In this study, a method of mid-level data fusion with the fingerprint region was proposed, which was combined with the characteristic wavelengths that contain fingerprint information in NIR and FT-MIR spectra to detect the DON level in FHB wheat during wheat processing. NIR and FT-MIR raw spectroscopy data on normal wheat and FHB wheat were obtained in the experiment. MSC was used for pretreatment, and characteristic wavelengths were extracted by CARS, MGS and XLW. The variables that can effectively reflect fingerprint information were retained to build the mid-level data fusion matrix. LS-SVM and PLS-DA were applied to investigate the performance of the single spectroscopic model, mid-level data fusion model and mid-level data fusion with fingerprint information model, respectively. The experimental results show that mid-level data fusion with a fingerprint information strategy based on fused NIR and FT-MIR spectra represents an effective method for the classification of DON levels in FHB wheat samples.

Spectral fusion modeling for soil organic carbon by a parallel input-convolutional neural network

Visible-to-near-infrared (vis–NIR) and mid-infrared (MIR) spectroscopy have been widely utilized for the quantitative estimation of soil organic carbon (SOC). The fusion of vis–NIR and MIR data can be hypothesized to provide accurate and reliable prediction for SOC because spectral data within a specific range of each individual sensor may lack important absorptive features associated with SOC. In this study, six data fusion strategies, principally direct concatenation-partial least squares regression (DC-PLSR), outer product analysis-PLSR (OPA-PLSR), OPA-competitive adaptive reweighted sampling-PLSR (OPA-CARS-PLSR), sequentially orthogonalized-PLSR (SO-PLSR), DC-convolutional neural network (DC-CNN), and parallel input-CNN (PI-CNN), were compared for the spectral estimations of SOC. The data fusion and individual sensor models were developed using soil samples collected from Zhejiang Province, East China, and scanned under laboratory conditions with both vis–NIR and MIR spectrophotometers. The validation results of vis–NIR (validation coefficient of determination [R2] = 0.63–0.73) were generally better than those of MIR (validation R2 = 0.45–0.59). For data fusion, the best validation accuracy was achieved by the PI-CNN (validation R2 = 0.84), followed in descending order by DC-CNN (validation R2 = 0.78), SO-PLSR (validation R2 = 0.73), OPA-CARS-PLSR (validation R2 = 0.69), OPA-PLSR (validation R2 = 0.66), and DC-PLSR (validation R2 = 0.64). The better performance of PI-CNN over DC-CNN demonstrates the necessity of using different sizes of convolutional kernels before feeding into the fully connected layers in the CNN network for fusing vis–NIR and MIR spectral data. The deep-learning fusion method based on PI-CNN can be considered an efficient tool for integrating data from multiple sensors for estimating soil properties in the field of soil spectral modeling.

A brown dwarf donor and an optically thin accretion disc with a complex stream impact region in the period-bouncer candidate BW Sculptoris

ABSTRACT We present an analysis of multi-epoch spectroscopic and photometric observations of the WZ Sge-type dwarf nova BW Scl, a period-bouncer candidate. We detected multiple irradiation-induced emission lines from the donor star allowing the radial velocity variations to be measured with high accuracy. Also, using the absorption lines Mg ii 4481 Å and Ca ii K originated in the photosphere of the accreting white dwarf (WD), we measured the radial velocity semi-amplitude of the WD and its gravitational redshift. We find that the WD has a mass of 0.85 ± 0.04 M⊙, while the donor is a low-mass object with a mass of 0.051 ± 0.006 M⊙, well below the hydrogen-burning limit. Using NIR data, we put an upper limit on the effective temperature of the donor to be ≲1600 K, corresponding to a brown dwarf of T spectral type. The optically thin accretion disc in BW Scl has a very low luminosity ≲4 × 1030 erg s−1 which corresponds to a very low-mass accretion rate of ≲7 × 10−13 M⊙ yr−1. The outer parts of the disc have a low density allowing the stream to flow down to the inner disc regions. The brightest part of the hotspot is located close to the circularization radius of the disc. The hotspot is optically thick and has a complex elongated structure. Based on the measured system parameters, we discuss the evolutionary status of the system.

Single-Oxidation of Ni(II) Complexes of Bidentate Fused N-Heterocyclic Carbene–Phenol Conjugates

The pro-ligands (H2LtBu)+ and (H2LOMe)+ were synthesized as chloride salts. They associate an imidazo [1,5-a] pyridinium ring and an electron-rich 2,4-(di-tert-butyl)phenol or 2-(tert-butyl)-4-methoxyphenol, respectively. The homoleptic Ni(II) complexes of H2LtBu (1) and H2LOMe (2) were prepared and structurally characterized, showing a four-coordinate nickel center with two ligands bound by one phenolate oxygen and one carbene each. Complexes 1 and 2 show two reversible oxidation waves in their cyclic voltammetry curve, at E1/21 = 0.19 V and E1/22 = 0.59 V, and E1/21 = 0.03 V and E1/22 = 0.32 V, respectively. The one-electron oxidized forms were generated and characterized by spectroscopy, while 2+ could be isolated as single crystals. They both demonstrate two remarkable NIR bands indicating radical species in CH2Cl2. The EPR parameters are 2.097, 2.023, and 1.997 (gaverage = 2.039) for 1+ and 2.030, 2.016, and 1.994 (gaverage = 2.013) for 2+. Both gaverage values are slightly higher than those reported for genuine phenoxyl radicals but far below those of Ni(III) complexes, supporting a metalloradical character of 1+ and 2+, whereby one metal d-orbital contributes to the singly occupied molecular orbital (SOMO). A very good agreement is found between DFT calculated and experimental vis–NIR and EPR data. Both metalloradical cations could be converted into Ni(III) species by addition of pyridine. The redox isomerism could be monitored by EPR spectroscopy: the [1•Py]+ species exhibits for instance the g-values 2.157, 2.136, and 2.014 (gaverage = 2.103). The Ni(III) adducts are thermally unstable and undergo reduction back to the neutral precursor at room temperature.

Low-value recyclable waste identification based on NIR feature analysis and RGB-NIR fusion

Low-value recyclable waste is one of the main components of solid waste, which is challenging to recycle because of its low density, small size, multiple categories, and large quantity. However, the primarily used spectral-based sorting is expensive, so incineration and landfill are still the main disposal methods in developing countries. In order to reduce the cost of sorting equipment, improve the accuracy and efficiency of identification, and facilitate the promotion of automatic sorting, this paper combines RGB and limited NIR data to realize cost-effective sorting. Firstly, an RGB-NIR instance segmentation dataset of white, color, and transparent PE, PP, and PET plastic flakes was produced, which are the main components of low-value recyclable waste. Then PCA and ANOVA were used to analyze the characteristics of the amplitude and band ratio of NIR data, where six characteristic bands were obtained. After that, the effect of the multi-channel NIR instance segmentation algorithm was tested based on 12 combinations of 6 bands, and the F1-score of the two better band combinations (1193, 1218, 1666 nm and 1134, 1193, 1218, 1349, 1398, 1666 nm) were 0.965 and 0.980, respectively. Finally, the selected 3-band NIR data and RGB data were mixed to create a fusion identification algorithm, which achieved the best material recognition accuracy (F1-score = 0.981) and color recognition accuracy (F1-score = 0.972). Moreover, the 9-categories comprehensive F1-score of the RGB-NIR fusion model is 0.961, which shows that it can recognize both material and color recognition.

Facial feature embedded CycleGAN for VIS–NIR translation

Visible and near-infrared (VIS–NIR) heterogeneous face recognition remains a challenging task due to distinctions between spectral components of two modalities and insufficiently pairwise VIS–NIR data. Inspired by the cycle-consistent generative adversarial network (CycleGAN), this paper proposes a facial feature embedded CycleGAN to translate between VIS and NIR face images, aiming to enable the distributions of translated (fake) images to be similar as those of true images. To learn the particular feature of NIR or VIS domain while preserving common facial representation between VIS and NIR domains, a facial feature extractor (FFE), tailored specifically for extracting effective feature from face images, is embedded in the generator of original CycleGAN. For implementing the FFE, we use the MobileFaceNet which is pre-trained on a VIS face database. The domain-invariant feature learning is enhanced by proposing a new pixel consistency loss. Additionally, we establish a new WHU VIS–NIR database including varies in face rotation and expressions to enrich the insufficient training data. Moreover, experiments on the well-known Oulu-CASIA NIR–VIS database and our WHU VIS–NIR database validate the potential benefit of the proposed FFE-based CycleGAN (FFE-CycleGAN). In particular, we achieve 96.5% accuracy on Oulu-CASIA and 98.9% accuracy on WHU VIS–NIR.

Insights and protocols for discrimination of sugarcane clones by dissimilarity measures on RGB and NIR data.

In sugarcane breeding, dense experiments have been considered in the initial phase (T1), such as the Simplified System (SS) of genotype selection. In this method, the seedlings of each family are transplanted directly from the seed box to the field, forming a kind of carpet. Despite the practical aspect of the method, selection problems are common, as stalks from the same individual within the family are subject to being taken to later evaluation stages, to the detriment of stalks from different individuals. To facilitate the discrimination of stalks of the same family in SS, we evaluated using RGB images (red:green:blue) and NIR (near infrared) spectra. We applied Euclidean distance (D) and Mahalanobis distance (D2) dissimilarity measures to the image and spectral data to distinguish stalks with different genotypes. RGB and NIR data were taken from type +1 leaf samples collected from two experimental blocks, totaling 31 evaluated families. The analyzes were carried out in two stages. In the first stage, we sought to evaluate the classification capacity using RGB images and NIR spectra, using D as a measure of dissimilarity. In the second step, we developed and validated a protocol using RGB images to classify clones, with D2 as a dissimilarity measure. Preliminary results, with distance D, allowed to discriminate clones based on the distance of the evaluated attributes and their combinations. In addition, with the analyzes using the D distance, it was identified that only the use of the R attribute (red band) would give satisfactory results for the second stage, which was the proposed analysis protocol, applying the D2 distance. The D2 statistic and associated p-value confirmed the protocol's usefulness in discriminating stalks in SS, especially stalks from the same families.

NIR quantitative model trans-scale calibration from small scale to pilot scale via directed DOSC-SBC algorithm

In order to solve the problem of inapplicability of NIR quantitative models due to the large difference between the modeling samples and the samples to be tested, Directed DOSC-SBC（DDOSC-SBC）algorithm is proposed in this paper based on Direct Orthogonal Signal Correction combined with Slope/Bias Correction (DOSC-SBC) algorithm. To obtain the suitable spectral matrix transfer parameters for the test set during DDOSC spectral preprocessing, several representative test samples in the test set were selected, then the spectral systematic errors between the modeling set and the test set were corrected with the SBC method in order to realize the trans-scale prediction of the NIR quantitative model. NIR data and the critical quality attributes（CQAs）were detected in the small scale and pilot scale pharmaceutical process of the fluidized bed granulation of dextrin and water extraction of honeysuckle. After the small scale model was calibrated via the directed DOSC-SBC algorithm which was guided by representative pilot scale samples, the small scale model was able to predict the pilot scale test samples more accurately. The NIR quantitative model trans-scale calibration from small scale to pilot scale was also successfully realized with a RPD value higher than 3.5 and RSEP value lower than 10%. DDOSC-SBC algorithm is a successful model trans-scale calibrated method that can be applied to NIR real-time monitoring of CQAs in the preparation process of Chinese herbal medicine.

Consensual Regression of Lasso-Sparse PLS models for Near-Infrared Spectra of Food

In some cases, near-infrared spectra (NIRS) make the prediction of quantitative models unreliable, and the choice of a suitable number of latent variables (LVs) for partial least square (PLS) is difficult. In this case, a strategy of fusing member models with important information is gradually becoming valued in recent research. In this work, a series of PLS regression models were developed with an increasing number of LVs as member models. Then, the least absolute shrinkage and selection operator (Lasso) was employed as the model’s selection access to sparse uninformative ones among these PLS member models. Deviation weighted fusion (DW-F), partial least squares regression coefficient fusion (PLS-F), and ridge regression coefficient fusion (RR-F) were comparatively used further to fuse the above sparsed member models, respectively. Three spectral datasets, including six attributes in NIR data of corn, apple, and marzipan, respectively, were applied in order to validate the feasibility of this fusion algorithm. Six fusion models of the above attributes performed better than the general optimal PLS model, with a noticeable enhancement of root mean errors squared of prediction (RMSEP) arriving at its highest at 80%. It also reduced more than half of the spectral bands; the DW-F especially showed its excellent fusing capacity and obtained the best performance. Results show that the preferred strategy of DW-F model combined with Lasso selection can make full use of spectral information, and significantly improve the prediction accuracy of fusion models.

Variable selection and data fusion for diesel cetane number prediction

This study evaluates the potential of variable selection to improve the performance of data fusion modelling to estimate diesel cetane number from NIR spectroscopy information acquired on total effluent samples obtained from the hydrocracking process and their operating variables. The evaluation conducted in this research was divided into four steps. First, predictive models were developed using each data block separately. Next, seven variable selection methods were applied on the NIR block, and eleven methods were applied on the process variable block. Then, with each data set generated from the variable selection analysis, single prediction models were generated and compared with those developed in the first step. Finally, data fusion was performed once the best variable selection method was defined for each data block. Two data fusion models were generated, a first using all the variables in the two blocks and a second using only the previously selected variables. In addition, the potential of the sequential and orthogonalized covariance selection (SO-CovSel) method was also analyzed. The results showed that the data fusion modelling using all variables from each data block improves the estimation of the diesel cetane number compared to single models (about 20% reduction of the RMSEP). However, using variable selection analysis before data fusion significantly improves the estimation of this property and leads to greater model stability regarding the RMSE's and r′s (about 47% of the RMSEP). The Covariance Selection (CovSel) method was the most efficient in the NIR data block, while for the process variable data block, it was the sequential backward floating feature selection method (SBFFS) that gave the best performance. The advantages offered by the variable selection resulted not only in having a more accurate prediction of the property but also in improving the analysis and understanding of the process by determining the variables that significantly impact the property studied.

Selecting relevant wavelength intervals for PLS calibration based on absorbance interquartile ranges

Wavelength selection is an important step in NIR data analysis as it yields more robust and accurate models for predicting products’ properties. This paper proposes a novel framework for wavelength selection in Partial Least Squares (PLS) models based on the absorbance interquartile range. The framework first divides samples into quartiles according to the response variable. The average quartile absorbance value is calculated for each wavelength; the distance in averages from samples belonging to the first and fourth quartiles is then used to generate a wavelength importance ranking, such that wavelengths with larger differences are ranked higher and expected to better explain variations in the response variable. The ranking guides the iterative inclusion of wavelength intervals into PLS models, starting with the interval presenting the largest average absorbance differences. The proposed framework was tested in three public datasets comprised of 10 response variables describing diesel, corn and soil properties, substantially reducing prediction error and the percentage of wavelengths included in the PLS models. When compared to four competing wavelength selection methods from the literature, the proposed framework yielded the best results in 7 out of 10 response variables in the training set, and in 6 out of 10 response variables in the testing set.

Prediction of Soil Available Boron Content in Visible-Near-Infrared Hyperspectral Based on Different Preprocessing Transformations and Characteristic Wavelengths Modeling.

The trace element boron (Boron, B) is an important factor in crops' development, pollination, and fertilization. Available boron (AB) in soil is the main source of boron nutrient absorption for crops. Rapid detection of AB is of great significance for crop nutrition diagnosis, soil testing and fertilization, precision agriculture development, scientific production management, and guarantee of stable yield and high quality. In this study, we propose a new method to predict soil available boron content using handheld nonimaging hyperspectroscopy in the visible-near-infrared range (350–1655 nm). As boron content is one of the fewest soil chemical elements, a rapid and accurate method has yet to be developed to detect and quantify the soil available boron. Visible-near-infrared ray (VIS-NIR) spectroscopy is widely utilized in the detection and quantification of soil available nutrients. There is, however, scant research on the detection of soil boron based on NIR data, and the performance of current regression model is still far from satisfactory. Our soil samples were collected from southern Anhui, China, with their NIR spectroscopy examined and the NIR data pretreated by 29 transformations and modeled with 10 regression algorithms. Of all the tested methods, SVM_RBF, BPNN, and PLS_RBF algorithms demonstrated the best performance and gave 0.80∼0.82 coefficient of determination value. At the same time, Random Forest algorithm (RFA), Successive Projection Algorithm (SPA), and Variable Importance in Projection (VIP) were used to extract the spectral characteristic wavelength data of soil available boron, and then the characteristic wavelength data were modeled with three regression algorithms: SVM_RBF, PLS_RBF, and BPNN. A comparative analysis of the prediction performance (R2, RPD, RMSE, and RPIQ) of the models established at the full band showed that the RFA-MSC/BPNN model achieved the best performance. Compared with the best full-wavelength model DT/SVM_RBF, the test set achieved a 3.06% increase in R2, a 7.12% drop in RMSE, a 7.71% gain in RPD, and a 7.78% increase in RPIQ. Our work sheds lights on how to achieve rapid quantification of the soil available boron concentration.

Unscrambling the Provenance of Eggs by Combining Chemometrics and Near-Infrared Reflectance Spectroscopy.

Issues related to food authenticity, traceability, and fraud have increased in recent decades as a consequence of the deliberate and intentional substitution, addition, tampering, or misrepresentation of food ingredients, where false or misleading statements are made about a product for economic gains. This study aimed to evaluate the ability of a portable NIR instrument to classify egg samples sourced from different provenances or production systems (e.g., cage and free-range) in Australia. Whole egg samples (n: 100) were purchased from local supermarkets where the label in each of the packages was used as identification of the layers’ feeding system as per the Australian legislation and standards. The spectra of the albumin and yolk were collected using a portable NIR spectrophotometer (950–1600 nm). Principal component analysis (PCA) and linear discriminant analysis (LDA) were used to analyze the NIR data. The results obtained in this study showed how the combination of chemometrics and NIR spectroscopy allowed for the classification of egg albumin and yolk samples according to the system of production (cage and free range). The proposed method is simple, fast, environmentally friendly and avoids laborious sample pre-treatment, and is expected to become an alternative to commonly used techniques for egg quality assessment.

Estimating pasture aboveground biomass under an integrated crop-livestock system based on spectral and texture measures derived from UAV images

Pasture production under integrated crop-livestock systems (ICLS) is a key element to support grazing management decisions. Therefore, further experimentation is required to produce reliable estimates of pasture productivity. An unmanned aerial vehicle (UAV) is a viable tool to obtain fast and accurate aboveground biomass (AGB) estimates in pastures under ICLS. We tested several datasets of variables composed of original spectral bands (RGB-NIR), vegetation indices, and gray-level cooccurrence matrix (GLCM) textures to estimate pasture AGB using the random forest (RF) algorithm and feature selection methods in a commercial ICLS farm in western São Paulo State, Brazil. Field measures of pasture AGB were carried out in three field campaigns over five months to capture the spatiotemporal variability of the pasture fields. Most tested models reached similar results on pasture AGB estimates (R2: 0.60 to 0.70), while the number of variables selected for the RF algorithm differed among models (from 6 to 160 variables). The three more accurate models used few variables, two of which used only texture measures, while the third also employed combined spectral bands and vegetation indices. The best model (R2 = 0.70) used only 10 texture measures. Texture measures that represented images’ inner patterns, calculated from NIR, red-edge, and triangular greenness index data, were the most relevant variables for the main models. The texture contribution indicates important information to be considered to estimate pasture AGB when using UAV imagery. The results achieved by UAV estimates allow the design of multitemporal AGB pasture maps, which may be useful for building more reliable spatiotemporal data.

NIR spectroscopy as an environmentally friendly method to determine the lipid and carbohydrate concentration in Lintong Arabica coffee bean

Proximate concentration of green bean coffee is known as an important quality which influence the final flavour of coffee beverage. Lipid and carbohydrate are two of the proximate concentration, so that determination of those concentration in green coffee bean is needed. Commonly, determination of lipid and carbohydrate concentration was conducted using chemical method. This method spend time and destructive, so it is not suitable for industries which need a fast, nondestructive and environmentally friendly measurement. This study aimed to calibrate the NIR and chemical data of lipid and carbohydrate concentration using PLS method to build a real time prediction of coffee bean concentration. This study used a NIR wave of 1000-2500 nm, followed by determination of lipid and carbohydrate concentration by proximate analysis, first and second derivative, multiple scatter correction and standard normal variate as NIR data pretreatment. Result showed that lipid and carbohydrate concentration of Lintong arabica coffee bean could be well predicted using FT-NIRS.

Orthogonal Modality Disentanglement and Representation Alignment Network for NIR-VIS Face Recognition

Near-infrared and visual (NIR-VIS) face matching, as the most typical task in Heterogeneous Face Recognition (HFR), has attracted increasing attention in recent years. However, due to the large within-class discrepancies, including domain differences and residual discrepancies (i.e., lighting, expressions, occlusion, blurry, pose, etc), this is still a difficult task. Conventional NIR-VIS FR methods only focus on reducing the modality gap between cross-domain images, while neglecting to eliminate the residual variations. To better solve the above problems, this paper proposes a novel Orthogonal Modality Disentanglement and Representation Alignment (OMDRA) approach, which consists of three key components, including Modality-Invariant (MI) loss, Orthogonal Modality Disentanglement (OMD) and Deep Representation Alignment (DRA). Firstly, the MI loss is designed to learn modality-invariant and identity-discriminative representation, by increasing between-class separability and within-class compactness between NIR and VIS heterogeneous data. Secondly, the high-level Hybrid Facial Feature (HFF) layer of the backbone network is projected into two subspaces: the modality-related and identity-related subspaces. The OMD is designed to decouple modal information via an adversarial process, and we further impose Orthogonal Representation Decorrelation (ORD) to the OMD to decrease the correlation between identity representations and domain representations, as well as enhancing their representation capabilities. Finally, the DRA aims to eliminate the residual variations by performing a high-level representation alignment between non-neutral face and neutral face, which can effectively guides the network to learn discriminative and residual-invariant face representation. The joint scheme enables the disentanglement of modality variations, elimination of residual discrepancies, and the purification of identity information. Extensive experiments on challenging cross-domain databases indicate that our OMDRA method is superior to the state-of-the-art methods.

Strategies for efficient estimation of soil organic content at the local scale based on a national spectral database

AbstractSoil function degradation threatens the sustainable management of soil resources and soil organic matter (SOM) is a vital and important factor. Powerful measuring tools will become very important, especially in areas where data are poor or absent. The archive: China Soil Visible and Near Infrared (vis–NIR) Spectroscopy Library (CSSL) could help providea solution for less costly and fast measuring of SOM. The aim of this article was to compare SOM prediction performance according to three strategies: i) general global partial least squares regression (PLSR) using CSSL with and without spiking samples; ii) memory‐based learning (MBL) using CSSL with and without spiking samples; and iii) general PLSR using only spiking samples to predict soil organic matter in the target area. When using spiked subsets, we also investigated the prediction performance of the extra‐weighted (several copies) subsets. A series of spiking subsets were randomly selected from the total spiking samples, which were selected by conditioned Latin hypercube sampling (cLHS) from the target sites. We calculated only the mean squared Euclidean distance (msd) between the estimates density function (pds) of the principal components (PCs) of vis–NIR spectroscopy from the validation dataset and spiking subsets and statistically inferred the optimal sampling set size to be 30. Our study showed that global PLSR using CSSL spiked with the statistically optimal local samples can achieve higher predicted performance [with a mean root mean square error (RMSE) of 5.75]. MBL spiked with five extra‐weighted optimal spiking samples achieved the best accuracy with an RMSE of 3.98, an R2 of 0.70, a bias of 0.04, and an LCCC of 0.81. The msd is a simple and effective method to determine an adequate spiking set size using only vis–NIR data. These accurate predictions demonstrated the usefulness of statistically representative spiking and MBL for advanced large soil spectral libraries for SOM determination, which is currently lacking at large soil spectral libraries in use.

Research and Application Validation of a Feature Wavelength Selection Method Based on Acousto-Optic Tunable Filter (AOTF) and Automatic Machine Learning (AutoML).

Near-infrared spectroscopy has been widely applied in various fields such as food analysis and agricultural testing. However, the conventional method of scanning the full spectrum of the sample and then invoking the model to analyze and predict results has a large amount of collected data, redundant information, slow acquisition speed, and high model complexity. This paper proposes a feature wavelength selection approach based on acousto-optical tunable filter (AOTF) spectroscopy and automatic machine learning (AutoML). Based on the programmable selection of sub nm center wavelengths achieved by the AOTF, it is capable of rapid acquisition of combinations of feature wavelengths of samples selected using AutoML algorithms, enabling the rapid output of target substance detection results in the field. The experimental setup was designed and application validation experiments were carried out to verify that the method could significantly reduce the number of NIR sampling points, increase the sampling speed, and improve the accuracy and predictability of NIR data models while simplifying the modelling process and broadening the application scenarios.

NIR-Based Intelligent Sensing of Product Yield Stress for High-Value Bioresorbable Polymer Processing.

PLA (polylactide) is a bioresorbable polymer used in implantable medical and drug delivery devices. Like other bioresorbable polymers, PLA needs to be processed carefully to avoid degradation. In this work we combine in-process temperature, pressure, and NIR spectroscopy measurements with multivariate regression methods for prediction of the mechanical strength of an extruded PLA product. The potential to use such a method as an intelligent sensor for real-time quality analysis is evaluated based on regulatory guidelines for the medical device industry. It is shown that for the predictions to be robust to processing at different times and to slight changes in the processing conditions, the fusion of both NIR and conventional process sensor data is required. Partial least squares (PLS), which is the established ’soft sensing’ method in the industry, performs the best of the linear methods but demonstrates poor reliability over the full range of processing conditions. Conversely, both random forest (RF) and support vector regression (SVR) show excellent performance for all criteria when used with a prior principal component (PC) dimension reduction step. While linear methods currently dominate for soft sensing of mixture concentrations in highly conservative, regulated industries such as the medical device industry, this work indicates that nonlinear methods may outperform them in the prediction of mechanical properties from complex physicochemical sensor data. The nonlinear methods show the potential to meet industrial standards for robustness, despite the relatively small amount of training data typically available in high-value material processing.

Benchtop NIR data standardization on handheld spectrometers to identify paracetamol tablets

AbstractNear‐infrared spectroscopy (NIRS) allows innovative applications in terms of quality control, for example, in raw materials verification, in process analytical technology (PAT), and in discrimination between genuine and falsified medicines. The development of small and cheap handheld devices is expanding in the field, while trying to keep similar performances as benchtop instruments have. Considering traceability and quality control of drug compounds, this work is intended to identify 13 different paracetamol tablets on the Belgian market by using NIRS. The performances of a Fourier‐transform near‐infrared spectroscopy (FT‐NIR) benchtop and two handheld NIR spectrometers were investigated comparatively. All spectra were collected through the blister in the reflectance diffuse mode. NIR spectral fingerprints were pretreated and analyzed using principal component analysis (PCA) and classified with soft independent modeling of class analogy (SIMCA). The performances of the spectrometers were evaluated after standardization of the reference benchtop database to the handheld spectrometers. The instrumental response of the benchtop spectrometer was standardized towards the handheld device using the piecewise direct standardization (PDS) algorithm. These investigations permitted the advantages and limitations of NIR data standardization on predictive models to be pointed out. The SIMCA models based on the spectral data of the benchtop (B), handheld 1 (H1), and handheld 2 (H2) instruments had an accuracy of 99.2%, 97.7%, and 96.2%, respectively. The same accuracy was obtained with the models of the transferred benchtop spectra for the H1 and H2 devices. However, adding handheld spectra to the transferred benchtop spectra resulted in models with improved accuracy: from 97.7% to 98.5% for the H1 instrument and from 96.2% to 97.7% for the H2 instrument. The proposed strategy highlights the potential of using a reference database collected by NIRS for pharmaceutical analyses by handheld NIR spectrometers in terms of traceability.