Functional data geometric morphometrics with machine learning for craniodental shape classification in shrews.

This work proposes a functional data analysis approach for morphometrics in classifying three shrew species (S. murinus, C. monticola, and C. malayana) from Peninsular Malaysia. Functional data geometric morphometrics (FDGM) for 2D landmark data is introduced and its performance is compared with classical geometric morphometrics (GM). The FDGM approach converts 2D landmark data into continuous curves, which are then represented as linear combinations of basis functions. The landmark data was obtained from 89 crania of shrew specimens based on three craniodental views (dorsal, jaw, and lateral). Principal component analysis and linear discriminant analysis were applied to both GM and FDGM methods to classify the three shrew species. This study also compared four machine learning approaches (naïve Bayes, support vector machine, random forest, and generalised linear model) using predicted PC scores obtained from both methods (a combination of all three craniodental views and individual views). The analyses favoured FDGM and the dorsal view was the best view for distinguishing the three species.

A quantitative test of the "Ecomorphotype Hypothesis" for fossil true seals (Family Phocidae).

The fossil record of true seals (Family Phocidae) is mostly made up of isolated bones, some of which are type specimens. Previous studies have sought to increase referral of non-overlapping and unrelated fossils to these taxa using the 'Ecomorphotype Hypothesis', which stipulates that certain differences in morphology between taxa represent adaptations to differing ecology. On this basis, bulk fossil material could be lumped to a specific ecomorphotype, and then referred to species in that ecomorphotype, even if they are different bones. This qualitative and subjective method has been used often to expand the taxonomy of fossil phocids, but has never been quantitatively tested. We test the proposed ecomorphotypes using morphometric analysis of fossil and extant northern true seal limb bones, specifically principal components analysis and discriminant function analysis. A large amount of morphological overlap between ecomorphotypes, and poor discrimination between them, suggests that the 'Ecomorphotype Hypothesis' is not a valid approach. Further, the analysis failed to assign fossils to ecomorphotypes designated in previous studies, with some fossils from the same taxa being designated as different ecomorphotypes. The failure of this approach suggests that all fossils referred using this method should be considered to have unknown taxonomic status. In light of this, and previous findings that phocid limb bones have limited utility as type specimens, we revise the status of named fossil phocid species. We conclude that the majority of named fossil phocid taxa should be considered nomina dubia.

Differentiation of Listeria monocytogenes serotypes using near infrared hyperspectral imaging.

Among the severe foodborne illnesses, listeriosis resulting from the pathogen Listeria monocytogenes exhibits one of the highest fatality rates. This study investigated the application of near infrared hyperspectral imaging (NIR-HSI) for the classification of three L. monocytogenes serotypes namely serotype 4b, 1/2a and 1/2c. The bacteria were cultured on Brain Heart Infusion agar, and NIR hyperspectral images were captured in the spectral range 900-2500 nm. Different pre-processing methods were applied to the raw spectra and principal component analysis was used for data exploration. Classification was achieved with partial least squares discriminant analysis (PLS-DA). The PLS-DA results revealed classification accuracies exceeding 80 % for all the bacterial serotypes for both training and test set data. Based on validation data, sensitivity values for L. monocytogenes serotype 4b, 1/2a and 1/2c were 0.69, 0.80 and 0.98, respectively when using full wavelength data. The reduced wavelength model had sensitivity values of 0.65, 0.85 and 0.98 for serotype 4b, 1/2a and 1/2c, respectively. The most relevant bands for serotype discrimination were identified to be around 1490 nm and 1580-1690 nm based on both principal component loadings and variable importance in projection scores. The outcomes of this study demonstrate the feasibility of utilizing NIR-HSI for detecting and classifying L. monocytogenes serotypes on growth media.

Evaluation of EEG Signals by Spectral Peak Methods and Statistical Correlation for Mental State Discrimination Induced by Arithmetic Tasks.

Bringing out brain activity through the interpretation of EEG signals is a challenging problem that involves combined methods of signal analysis. The issue of classifying mental states induced by arithmetic tasks can be solved through various classification methods, using diverse characteristic parameters of EEG signals in the time, frequency, and statistical domains. This paper explores the results of an experiment that aimed to highlight arithmetic mental tasks contained in the PhysioNet database, performed on a group of 36 subjects. The majority of publications on this topic deal with machine learning (ML)-based classification methods with supervised learning support vector machine (SVM) algorithms, K-Nearest Neighbor (KNN), Linear Discriminant Analysis (LDA), and Decision Trees (DTs). Also, there are frequent approaches based on the analysis of EEG data as time series and their classification with Recurrent Neural Networks (RNNs), as well as with improved algorithms such as Long Short-Term Memory (LSTM), Bidirectional Long Short-Term Memory (BLSTM), and Gated Recurrent Units (GRUs). In the present work, we evaluate the classification method based on the comparison of domain limits for two specific characteristics of EEG signals: the statistical correlation of pairs of signals and the size of the spectral peak detected in theta, alpha, and beta bands. This study provides some interpretations regarding the electrical activity of the brain, consolidating and complementing the results of similar research. The classification method used is simple and easy to apply and interpret. The analysis of EEG data showed that the theta and beta frequency bands were the only discriminators between the relaxation and arithmetic calculation states. Notably, the F7 signal, which used the spectral peak criterion, achieved the best classification accuracy (100%) in both theta and beta bands for the subjects with the best results in performing calculations. Also, our study found the Fz signal to be a good sensor in the theta band for mental task discrimination for all subjects in the group with 90% accuracy.

Evaluation of a Voltametric E-Tongue Combined with Data Preprocessing for Fast and Effective Machine Learning-Based Classification of Tomato Purées by Cultivar.

The potential of a voltametric E-tongue coupled with a custom data pre-processing stage to improve the performance of machine learning techniques for rapid discrimination of tomato purées between cultivars of different economic value has been investigated. To this aim, a sensor array with screen-printed carbon electrodes modified with gold nanoparticles (GNP), copper nanoparticles (CNP) and bulk gold subsequently modified with poly(3,4-ethylenedioxythiophene) (PEDOT), was developed to acquire data to be transformed by a custom pre-processing pipeline and then processed by a set of commonly used classifiers. The GNP and CNP-modified electrodes, selected based on their sensitivity to soluble monosaccharides, demonstrated good ability in discriminating samples of different cultivars. Among the different data analysis methods tested, Linear Discriminant Analysis (LDA) proved to be particularly suitable, obtaining an average F1 score of 99.26%. The pre-processing stage was beneficial in reducing the number of input features, decreasing the computational cost, i.e., the number of computing operations to be performed, of the entire method and aiding future cost-efficient hardware implementation. These findings proved that coupling the multi-sensing platform featuring properly modified sensors with the custom pre-processing method developed and LDA provided an optimal tradeoff between analytical problem solving and reliable chemical information, as well as accuracy and computational complexity. These results can be preliminary to the design of hardware solutions that could be embedded into low-cost portable devices.

A supramolecular fluorescence sensor array for the differentiation of multiple anions and prediction of iodine in artificial urine using machine learning.

The simultaneous discrimination and detection of multiple anions in an aqueous solution has been a major challenge due to their structural similarity and low charge radii. In this study, we have constructed a supramolecular fluorescence sensor array based on three host-guest complexes to distinguish five anions (F-, Cl-, Br-, I-, and ClO-) in an aqueous solution using anionic-induced fluorescence quenching combined with linear discriminant analysis. Due to the different affinities of the three host-guest complexes for each anion the anion quenching efficiency for each host-guest complex was likewise different, and the five anions were well recognized. The fluorescence sensor array not only distinguished anions at different concentrations (0.5, 10, and 50 µM) with 100% accuracy but also showed good linearity within a certain concentration range. The limit of detection (LOD) was < 0.5 µM. Our interference study showed that the developed sensor array had good anti-interference ability. The practicability of the developed sensor array was also verified by the identification and differentiation of toothpaste brands with different fluoride content and the prediction of the iodine concentration in urine combined with machine learning.

Differential effects of mass-loading the eardrum and stiffening the middle ear on wideband absorbance.

The current work investigated the effects of mass-loading the eardrum on wideband absorbance in humans. A non-invasive approach to mass-loading the eardrum was utilized in which water was placed on the eardrum via ear canal access. The mass-loaded absorbance was compared to absorbance measured for two alternative middle ear states: normal and stiffened. To stiffen the ear, subjects pressurized the middle ear through either exsufflation or insufflation concurrent with Eustachian tube opening. Mass-loading the eardrum was hypothesized to reduce high-frequency absorbance, whereas pressurizing the middle ear was hypothesized to reduce low- to mid-frequency absorbance. Discriminant linear analysis classification was performed to evaluate the utility of absorbance in differentiating between conditions. Water on the eardrum reduced absorbance over the 0.7- to 6-kHz frequency range and increased absorbance at frequencies below approximately 0.5 kHz; these changes approximated the pattern of changes reported in both hearing thresholds and stapes motion upon mass-loading the eardrum. Pressurizing the middle ear reduced the absorbance over the 0.125- to 4-kHz frequency range. Several classification models based on the absorbance in two- or three-frequency bands had accuracy exceeding 88%.

A Dynamic Window Method Based on Reinforcement Learning for SSVEP Recognition.

Steady-state visual evoked potential (SSVEP) is one of the most used brain-computer interface (BCI) paradigms. Conventional methods analyze SSVEPs at a fixed window length. Compared with these methods, dynamic window methods can achieve a higher information transfer rate (ITR) by selecting an appropriate window length. These methods dynamically evaluate the credibility of the result by linear discriminant analysis (LDA) or Bayesian estimation and extend the window length until credible results are obtained. However, the hypotheses introduced by LDA and Bayesian estimation may not align with the collected real-world SSVEPs, which leads to an inappropriate window length. To address the issue, we propose a novel dynamic window method based on reinforcement learning (RL). The proposed method optimizes the decision of whether to extend the window length based on the impact of decisions on the ITR, without additional hypotheses. The decision model can automatically learn a strategy that maximizes the ITR through trial and error. In addition, compared with traditional methods that manually extract features, the proposed method uses neural networks to automatically extract features for the dynamic selection of window length. Therefore, the proposed method can more accurately decide whether to extend the window length and select an appropriate window length. To verify the performance, we compared the novel method with other dynamic window methods on two public SSVEP datasets. The experimental results demonstrate that the novel method achieves the highest performance by using RL.

Identifying the origin of Yemeni green coffee beans using near infrared spectroscopy: a promising tool for traceability and sustainability.

Yemeni smallholder coffee farmers face several challenges, including the ongoing civil conflict, limited rainfall levels for irrigation, and a lack of post-harvest processing infrastructure. Decades of political instability have affected the quality, accessibility, and reputation of Yemeni coffee beans. Despite these challenges, Yemeni coffee is highly valued for its unique flavor profile and is considered one of the most valuable coffees in the world. Due to its exclusive nature and perceived value, it is also a prime target for food fraud and adulteration. This is the first study to identify the potential of Near Infrared Spectroscopy and chemometrics-more specifically, the discriminant analysis (PCA-LDA)-as a promising, fast, and cost-effective tool for the traceability of Yemeni coffee and sustainability of the Yemeni coffee sector. The NIR spectral signatures of whole green coffee beans from Yemeni regions (n = 124; Al Mahwit, Dhamar, Ibb, Sa'dah, and Sana'a) and other origins (n = 97) were discriminated with accuracy, sensitivity, and specificity ≥ 98% using PCA-LDA models. These results show that the chemical composition of green coffee and other factors captured on the spectral signatures can influence the discrimination of the geographical origin, a crucial component of coffee valuation in the international markets.

Discriminant analysis of ecological factors influencing sarcopenia in older people in South Korea.

This study aimed to investigate the ecological system factors that influence discrimination of sarcopenia among older individuals living in contemporary society. Data analysis included information from 618 older adults individuals aged 65 years or older residing in South Korea. To assess variations in ecological system factors related to SARC-F scores, we conducted correlation analysis and t-tests. Discriminant analysis was used to identify factors contributing to group discrimination. The key findings are summarized as follows. First, significant differences at the p < 0.001 level were observed between the SARC-F score groups in various aspects, including attitudes toward life, wisdom in life, health management, social support, media availability, sports environment, collectivist values, and values associated with death. Further, service environment differences were significant at p < 0.01 level, while social belonging and social activities exhibited significance at p < 0.05. Second, factors influencing group discrimination based on the SARC-F scores were ranked in the following order: health management, attitudes toward life, fear of own death, wisdom in life, physical environment, sports environment, media availability, social support, fear of the own dying, collectivist values, service environment, social activities, and social belonging. Notably, the SARC-F tool, which is used for sarcopenia discrimination, primarily concentrates on physical functioning and demonstrates relatively low sensitivity. Therefore, to enhance the precision of sarcopenia discrimination within a score-based group discrimination process, it is imperative to incorporate ecological system factors that exert a significant influence. These modifications aimed to enhance the clarity and precision of the text in an academic context.

Growth period determination and color coordinates visual analysis of tomato using hyperspectral imaging technology.

Growth period determination and color coordinates prediction are essential for comparing postharvest fruit quality. This paper proposes a tomato growth period judgment and color coordinates prediction model based on hyperspectral imaging technology. It utilizes the most effective color coordinates prediction model to obtain a color visual image. Firstly, hyperspectral images were taken of tomatoes at different growth periods (green-ripe, color-changing, half-ripe, and full-ripe), and color coordinates (L*, a*, b*, c, h) were obtained using a colorimeter. The sample set was divided by the sample set partitioning based on joint X-Y distances (SPXY). The support vector machine (SVM), K-nearest neighbors (KNN), and linear discriminant analysis (LDA) were used to discriminate growth period. Results show that the LDA model has the best prediction effect with a prediction set accuracy of 93.1%. In addition, effective wavelengths were selected using competitive adaptive reweighted sampling (CARS) and successive projections algorithm (SPA), and chromaticity prediction models were established using partial least squares regression (PLSR), multiple linear regression (MLR), principal component regression (PCR) and support vector machine regression (SVR) Finally, the color of each pixel of the tomato is calculated using the optimal model, generating a visual distribution image of the color coordinate. The results showed that hyperspectral imaging can non-destructively detect tomatoes' growth stage and color coordinates, providing great significance for designing a tomato quality grading system.

Discrimination of internal crack for rice seeds using near infrared spectroscopy.

It is an important thing to identify internal crack in seeds from normal seeds for evaluating the quality of rice seeds (Oryza sativa L.). In this study, non-destructive discrimination of internal crack in rice seeds using near infrared spectroscopy and chemometrics is proposed. Principal component analysis (PCA) was used to analyze the rice seeds spectra. Four supervised classification techniques(partial least squares discriminate analysis (PLS-DA), support vector machines (SVM), k-nearest neighbors (KNN) and random forest (RF)) with four different pre-processing techniques (standard normal variate (SNV), multiplicative scatter correction (MSC), first and second derivative with Savitzky-Golay (SG) smoothing) were applied. The best results (Sn = 0.8824, Sp = 0.9429, Acc = 0.913) were achieved by PLS-DA with the raw spectral data. The performance of the best SVM model was inferior to that of PLS-DA, but superior to that of RF and KNN. Except for PLS-DA, four different preprocessing techniques were improved the performance of the developed models. The important variables for discriminating internal cracks in rice seeds were related to the amylose. Overall, the all results demonstrated the feasibility of non-destructive discrimination of internal crack for rice seeds (Oryza sativa L.) using near infrared spectroscopy and chemometrics.

Multispectral imaging for distinguishing hybrid forest seeds of Corymbia spp. and Eucalyptus spp. from their progenitors.

In the forest industry, interspecific hybridization, such as Eucalyptus urograndis (Eucalyptus grandis × Eucalyptus urophylla) and Corymbia maculata × Corymbia torelliana, has led to the development of high-performing F1 generations. The successful breeding of these hybrids relies on verifying progenitor origins and confirming post-crossing, but conventional genotype identification methods are resource-intensive and result in seed destruction. As an alternative, multispectral imaging analysis has emerged as an efficient and non-destructive tool for seed phenotyping. This approach has demonstrated success in various crop seeds. However, identifying seed species in the context of forest seeds presents unique challenges due to their natural phenotypic variability and the striking resemblance between different species. This study evaluates the efficacy of spectral imaging analysis in distinguishing hybrid seeds of E. urograndis and C. maculata × C. torelliana from their progenitors. Four experiments were conducted: one for Corymbia spp. seeds, one for each Eucalyptus spp. batch separately, and one for pooled batches. Multispectral images were acquired at 19 wavelengths within the spectral range of 365-970 nm. Classification models based on Linear Discriminant Analysis (LDA), Random Forest (RF), and Support Vector Machine (SVM) was created using reflectance and reflectance features, combined with color, shape, and texture features, as well as nCDA transformed features. The LDA algorithm, combining all features, provided the highest accuracy, reaching 98.15% for Corymbia spp., and 92.75%, 85.38, and 86.00 for Eucalyptus batch one, two, and pooled batches, respectively. The study demonstrated the effectiveness of multispectral imaging in distinguishing hybrid seeds of Eucalyptus and Corymbia species. The seeds' spectral signature played a key role in this differentiation. This technology holds great potential for non-invasively classifying forest seeds in breeding programs.

Identifying Meat from Grazing or Feedlot Yaks Using Visible and Near-infrared Spectroscopy with Chemometrics.

The quality of meat can differ between grazing and feedlot yaks. The present study examined whether spectral fingerprints by visible and near-infrared (Vis-NIR) spectroscopy and chemo-metrics could be employed to identify the meat of grazing and feedlot yaks. Thirty-six 3.5-year-old castrated male yaks (164 ± 8.38 kg) were divided into grazing and feedlot yaks. After 5 months on treatment, liveweight, carcass weight, and dressing percentage were greater in the feedlot than in grazing yaks. The grazing yaks had greater protein content but lesser fat content than feedlot yaks. Principal component analysis (PCA) was able to identify the meat of the two groups to a great extent. Using either partial least squares discriminant analysis (PLS-DA) or the soft independent modeling of class analogies (SIMCA) classification, the meat could be differentiated between the groups. Both the original and processed spectral data had a high discrimination percentage, especially the PLS-DA classification algorithm, with 100% discrimination in the 400-2500 nm band. The spectral preprocessing methods can improve the discrimination percentage, especially for the SIMCA classification. It was concluded that the method can be employed to identify meat from grazing or feedlot yaks. The unerring consistency across different wavelengths and data treatments highlights the model's robustness and the potential use of NIR spectroscopy combined with chemometric techniques for meat classification. PLS-DA's accurate classification model is crucial for the unique evaluation of yak meat in the meat industry, ensuring product traceability and meeting consumer expectations for the authenticity and quality of yak meat raised in different ways.

Reassociation of Skeletal Remains Using Laser-Induced Breakdown Spectroscopy.

A major challenge in forensic anthropology and bioarcheology is the development of fast and effective methods for sorting commingled remains. This study assesses how portable laser-induced breakdown spectroscopy (LIBS) can be used to group skeletal remains based on their elemental profiles. LIBS spectra were acquired from the remains of 45 modern skeletons, with a total data set of 8388 profiles from 1284 bones. Spectral feature selection was conducted to reduce the spectral profiles to the peaks exhibiting the highest variation among individuals. Emission lines corresponding to 9 elements (Ca, P, C, K, Mg, Na, Al, Ba, and Sr) were found important for classification. Linear discriminant analysis (LDA) was concurrently used to classify each spectral profile. From the 45 individuals, each LIBS spectrum was successfully sorted to its corresponding skeleton with an average accuracy of 87%. These findings indicate that variation exists among the LIBS profiles of individuals' skeletal remains, highlighting the potential for portable LIBS technology to aid in the sorting of commingled remains.

Heterogeneous Analysis of Extracellular Vesicles for Osteosarcoma Diagnosis.

Osteosarcoma (OS) is the most prevalent primary tumor of bones, often diagnosed late with a poor prognosis. Currently, few effective biomarkers or diagnostic methods have been developed for early OS detection with high confidence, especially for metastatic OS. Tumor-derived extracellular vesicles (EVs) are emerging as promising biomarkers for early cancer diagnosis through liquid biopsy. Here, we report a plasmonic imaging-based biosensing technique, termed subpopulation protein analysis by single EV counting (SPASEC), for size-dependent EV subpopulation analysis. In our SPASEC platform, EVs are accurately sized and counted on plasmonic sensor chips coated with OS-specific antibodies. Subsequently, EVs are categorized into distinct subpopulations based on their sizes, and the membrane proteins of each size-dependent subpopulation are profiled. We measured the heterogeneous expression levels of the EV markers (CD63, BMP2, GD2, and N-cadherin) in each of the EV subsets from both OS cell lines and clinical plasma samples. Using the linear discriminant analysis (LDA) model, the combination of four markers is applied to classify the healthy donors (n = 37), nonmetastatic OS patients (n = 13), and metastatic patients (n = 12) with an area under the curve of 0.95, 0.92, and 0.99, respectively. SPASEC provides accurate EV sensing technology for early OS diagnosis.

Beef muscle discrimination based on two-trace two-dimensional correlation spectroscopy (2T2D COS) combined with snapshot visible-near infrared multispectral imaging.

The purpose of this work was to assess the potential of 2T2D COS PLS-DA (two-trace two-dimensional correlation spectroscopy and partial least squares discriminant analysis) in conjunction with Visible Near infrared multispectral imaging (MSI) as a quick, non-destructive, and precise technique for classifying three beef muscles -Longissimus thoracis, Semimembranosus, and Biceps femoris- obtained from three breeds - the Blonde d'Aquitaine, Limousine, and Aberdeen Angus. The experiment was performed on 240 muscle samples. Before performing PLS-DA, spectra were extracted from MSI images and processed by SNV (Standard Normal Variate), MSC (Multivariate Scattering Correction) or AREA (area under curve equal 1) and converted in synchronous and asynchronous 2T2D COS maps. The results of the study highlighted that combining synchronous and asynchronous 2T2D COS maps before performing PLS-DA was the best strategy to discriminate between the three muscles (100% of classification accuracy and 0% of error).

Geographical origin identification of Khao Dawk Mali 105 rice using combination of FT-NIR spectroscopy and machine learning algorithms.

The mislabelled Khao Dawk Mali 105 rice coming from other geographical region outside the Thung Kula Rong Hai region is extremely profitable and difficult to detect; to prevent retail fraud (that adversely affects both the food industry and consumers), it is vital to identify geographical origin. Near infrared spectroscopy can be used to detect the specific content of organic moieties in agricultural and food products. The present study implemented the combinatorial method of FT-NIR spectroscopy with chemometrics to identify geographical origin of Khao Dawk Mali 105 rice. Rice samples were collected from 2 different region including the north and northeast of Thailand. NIR spectra data were collected in range of 12,500 - 4,000 cm-1 (800-2,500 nm). Five machine learning algorithms including linear discriminant analysis (LDA), partial least squares discriminant analysis (PLS-DA), C-support vector classification (C-SVC), backpropagation neural networks (BPNN), hybrid principal component analysis-neural network (PC-NN) and K-nearest neighbors (KNN) were employed to classify NIR data of rice samples with full wavelength and selected wavelength by Extremely Randomized Trees (Extra trees) algorithm. Based on the findings, geographical origin of rice could be specified quickly, cheaply, and reliably using combination of NIRS and machine learning. All models creating by full wavelength and selected wavelength exhibited accuracy between 65 and 100 % for identifying geographical region of rice. It was proven that NIR spectroscopy may be used for the quick and non-destructive identification of geographical origin of Khao Dawk Mali 105 rice.

Applying virtual sample generation and ensemble modeling for improving the spectral diagnosis of cancer.

Cancer diagnosis plays a key role in facilitating treatment and improving survival rates of patients. The combination of near-infrared (NIR) spectroscopy with data-driven algorithms offers a rapid and cost-effective approach for such a task. Due to the limitations of objective cases, the number of tumor samples is usually smaller, and the resulting dataset exhibit the issues of class imbalance, which has a more serious impact on the performance of diagnostic models. To deal with class imbalance and improve the sensitivity, this work investigates the feasibility of NIR spectroscopy combined with virtual sample generation (VSG) as well as ensemble strategy for developing diagnostic models. Based on preliminary experiment, several learning algorithms such as discriminant analysis (DA) and partial least square-discriminant analysis (PLS-DA) are screened out as algorithms for constructing prediction models. Three algorithms of VSG including synthetic minority oversampling technique (SMOTE), Borderline-SMOTE and adaptive synthetic sampling (ADASYN) are used for experiment. A fixed sample subset composed of 27 cancer samples and 54 normal samples are hold out as the test set. Three training sets containing 5, 10, 25 minority class samples and 54 majority class samples are used for model development. The experimental result indicates that overall, with PLS-DA algorithm, all VSG approaches can significantly improve the sensitivity of cancer diagnosis for all cases of training sets with different minority samples, but ADASYN performs the best. It reveals that the integration of NIR, PLS-DA, and ADASYN is a promising tool package for developing diagnosis methods.

Rapid identification and determination of adulteration in medicinal Arnebiae Radix by combining near infrared spectroscopy with chemometrics.

The medicinal Arnebia Radix (AR) is one of widely-used Chinese herbal medicines (CHMs), usually adulterated with non-medicinal species that seriously compromise the quality of AR and affect patients' health. Detection of these adulterants is usually performed by using expensive and time-consuming analytical instruments. In this study, a rapid, non-destructive, and effective method was proposed to identify and determine the adulteration in the medicinal AR by near-infrared (NIR) spectroscopy coupled with chemometrics. 37 batches of medicinal AR samples originated from Arnebia euchroma (Royle) Johnst., 11 batches of non-medicinal AR samples including Onosma paniculatum Bur. et Franch and Arnebia benthamii (Wall. ex G. Don) Johnston, and 72 batches of adulterated AR samples were characterized by NIR spectroscopy. The data driven-soft independent modeling by class analogy (DD-SIMCA) and partial least squares-discriminant analysis (PLS-DA) were separately used to differentiate the authentic from adulterated AR samples. Then the PLS and support vector machine (SVM) were applied to predict the concentration of the adulteration in the adulterated AR samples, respectively. As a result, the classification accuracies of DD-SIMCA and PLS-DA models were 100% for the calibration set, and 96.7% vs. 100% for the prediction set. Moreover, the relative prediction deviation (RPD) values of PLS models reached 11.38 and 7.75 for quantifying two adulterants species, which were obviously superior to the SVM models. It can be concluded that the NIR spectroscopy coupled with chemometrics is feasible to identify the authentic from adulterated AR samples and quantify the adulteration in adulterated AR samples.