Ultraviolet-visible diffuse reflectance spectroscopy combined with chemometrics for rapid discrimination of Angelicae Sinensis Radix from its four similar herbs

Rapid and accurate quality evaluation of Angelicae Sinensis Radix based on near-infrared spectroscopy and Bayesian optimized LSTM network

The authentic traditional Chinese medicines (TCMs) including Angelicae Sinensis Radix (ASR) are the representative of high-quality herbals in China. However, ASR from authentic region being adulterated or counterfeited is frequently occurring, and there is still a lack of rapid quality evaluation methods for identifying the authentic ASR. In this study, the color features of ASR were firstly characterized. The results showed that the authentic ASR cannot be fully identified by color characteristics. Then near-infrared (NIR) spectroscopy combined with Bayesian optimized long short-term memory (BO-LSTM) was used to evaluate the quality of ASR, and the performance of BO-LSTM with common classification and regression algorithms was compared. The results revealed that following the pretreatment of NIR spectra, the optimal NIR spectra combined with BO-LSTM not only successfully distinguished authentic, non-authentic, and adulterated ASR with 100 % accuracy, but also accurately predicted the adulteration concentration of authentic ASR (R2 > 0.99). Moreover, BO-LSTM demonstrated excellent performance in classification and regression compared with common algorithms (ANN, SVM, PLSR, etc.). Overall, the proposed strategy could quickly and accurately evaluate the quality of ASR, which provided a reference for other TCMs.

Application of signal processing techniques for the spectroscopic analysis of dolutegravir and lamivudine: a comparative assessment and greenness appraisal

HIV treatment has greatly improved over the years, with the introduction of antiretroviral drugs that target the virus and suppress its replication. Dolutegravir and lamivudine are two such antiretroviral drugs that are commonly used in HIV treatment regimens. Herein, three spectrophotometric methods manipulating ratio spectra were developed for the simultaneous analysis of dolutegravir and lamivudine in their binary mixtures. These methods include mathematical processing stages like ratio difference method or signal processing approaches such as the first derivative of the ratio spectra, and continuous wavelet transform. The developed spectrophotometric methods exploit the characteristic spectral differences between dolutegravir and lamivudine in order to quantify them simultaneously. These methods have shown promising results in terms of sensitivity and selectivity as validated per the ICH guidelines. Moreover, these methods offer a straightforward and economical alternative to more intricate analytical methodologies like high-performance liquid chromatography. By incorporating the analytical eco-scale and AGREE for greenness evaluation of the proposed methods, we can further ensure that these techniques are effective and environmentally friendly, aligning with the principles of green chemistry. This evaluation will provide a comprehensive understanding of the environmental friendliness of these spectrophotometric methods in pharmaceutical analysis.

Distinction and quantification of Panax polysaccharide extracts via attenuated total reflectance-Fourier transform infrared spectroscopy with first-order derivative processing

Derivative spectroscopy is used to separate the small absorption peaks superimposed on the main absorption band, which is widely adopted in modern spectral analysis to increase both the valid spectral information and the identification accuracy. In this study, a method based on attenuated total reflectance-Fourier transform infrared spectroscopy (ATR-FTIR) with first-order derivative (FD) processing combined with chemometrics is proposed for rapid qualitative and quantitative analysis of Panax ginseng polysaccharides (PGP), Panax notoginseng polysaccharides (PNP), and Panax quinquefolius polysaccharides (PQP). First, ATR-FTIR with FD processing was used to establish the discriminant model combined with principal component analysis (PCA), partial least squares discriminant analysis (PLS-DA) and linear discriminant analysis (LDA). After that, two-dimensional ATR-FTIR based on single-characteristic temperature as external interference (2D-sATR-FTIR) was established using ATR-FTIR with FD processing. Then, ATR-FTIR with FD processing was combined with PLS to establish and optimize the quantitative regression model. Finally, the established discriminant model and 2D-sATR-FTIR successfully distinguished PGP, PNP and PQP, and the optimal PLS regression model had a good prediction ability for the Panax polysaccharide extracts content. This strategy provides an efficient, economical and nondestructive method for the distinction and quantification of PGP, PNP and PQP in a short detection time.

Novel fluorescence spectroscopy method coupled with N-PLS-R and PLS-DA models for the quantification of cannabinoids and the classification of cannabis cultivars

INTRODUCTION

Cannabis sativa L. inflorescences are rich in secondary metabolites, particularly cannabinoids. The most common techniques for elucidating cannabinoid composition are expensive technologies, such as high-pressure liquid chromatography (HPLC).

OBJECTIVES

We aimed to develop and evaluate the performance of a novel fluorescence spectroscopy-based method coupled with N-way partial least squares regression (N-PLS-R) and partial least squares discriminant analysis (PLS-DA) models to replace the expensive chromatographic methods for preharvest cannabinoid quantification.

METHODOLOGY

Fresh medicinal cannabis inflorescences were collected and ethanol extracts were prepared. Their excitation-emission spectra were measured using fluorescence spectroscopy and their cannabinoid contents were determined by HPLC-PDA. Subsequently, N-PLS-R and PLS-DA models were applied to the excitation-emission matrices (EEMs) for cannabinoid concentration prediction and cultivar classification, respectively.

RESULTS

The N-PLS-R model was based on a set of EEMs (n = 82) and provided good to excellent quantification of (-)-Δ9-trans-tetrahydrocannabinolic acid, cannabidiolic acid, cannabigerolic acid, cannabichromenic acid, and (-)-Δ9-trans-tetrahydrocannabinol (R² _CV and R² _pred  > 0.75; RPD > 2.3 and RPIQ > 3.5; RMSECV/RMSEC ratio < 1.4). The PLS-DA model enabled a clear distinction between the four major classes studied (sensitivity, specificity, and accuracy of the prediction sets were all ≥0.9).

CONCLUSIONS

The fluorescence spectral region (excitation 220-400 nm, emission 280-550 nm) harbors sufficient information for accurate prediction of cannabinoid contents and accurate classification using a relatively small data set.

Rapid quantification of goat milk adulteration with cow milk using Raman spectroscopy and chemometrics

As goat milk has a higher economic value compared to cow milk, the phenomenon of adulterating goat milk with cow milk appears in the market. In this study, the potential of Raman spectroscopy along with chemometrics was investigated for the authentication and quantitation of liquid goat milk adulterated with cow milk. First, the results of principal component analysis (PCA) showed that there were differences between the Raman spectra of cow and goat milk, which made quantitative experiments possible. For quantification, three different brands of cow milk and goat milk were selected randomly and adulterated goat milk with cow milk at the proportion of 5-95%. 342 samples were used for the construction of the partial least squares regression (PLSR) model with 80% for the calibration set and 20% for the prediction set. The PLSR model showed excellent performance in quantifying the level of adulteration, for the prediction set, with a coefficient of determination (R²) of 0.9781, root mean square error (RMSE) of 3.82%, and a ratio of prediction to deviation (RPD) of 6.8. The results demonstrated the potential of Raman spectroscopy as a rapid, low cost and non-destructive analytical tool for detecting adulteration in goat milk.

Use of near-infrared spectroscopy for the classification of medicinal cannabis cultivars and the prediction of their cannabinoid and terpene contents

Cannabis sativa L. is used to treat a wide variety of medical conditions, in light of its beneficial pharmacological properties of its cannabinoids and terpenes. At present, the quantitative chemical analysis of these active compounds is achieved through the use of laborious, expensive, and time-consuming technologies, such as high-pressure liquid-chromatography- photodiode arrays, mass spectrometer detectors (HPLC-PDA or MS), or gas chromatography-mass spectroscopy (GC-MS). Hence, we aimed to develop a simple, accurate, fast, and cheap technique for the quantification of major cannabinoids and terpenes using Fourier transform near infra-red spectroscopy (FT-NIRS). FT-NIRS was coupled with multivariate classification and regression models, namely partial least square-discriminant analysis (PLS-DA) and partial least squares regression (PLS-R) models. The PLS-DA model yielded an absolute major class separation (high-THC, high-CBD, hybrid, and high-CBG) and perfect class prediction. Using only three latent variables (LVs), the cross-validation and prediction model errors indicated a low probability of over-fitting the data. In addition, the PLS-DA model enabled the classification of chemovars with genetic-chemical similarities. The classification of high-THCA chemovars was more sensitive and more specific than the classifications of the remaining chemovars. The prediction of cannabinoid and terpene concentrations by PLS-R yielded 11 robust models with high predictive capabilities (R2CV and R2pred > 0.8, RPD >2.5 and RPIQ >3, RMSECV/RMSEC ratio <1.2) and additional 15 models whose performance was acceptable for initial screening purposes (R2CV > 0.7 and R2pred < 0.8, RPD >2 and RPIQ <3, 1.2 < RMSECV/RMSEC ratio <2). Our results confirm that there is sufficient information in the FT-NIRS to develop cannabinoid and terpene prediction models and major-cultivar classification models.

Spectroscopic profiling-based geographic herb identification by neural network with random weights

Daodi medicinal material plays an important role in traditional Chinese medicine (TCM). This study researches and validates the NNRW (neural network with random weights) model on spectroscopic profiling data for geographical origin identification. NNRW is a special neural network model that does not require an iterative training process. It has been proved effective in various resource-limited data-driven applications. However, whether NNRW works for spectroscopic profiling data remains to be explored. In this study, the Raman and UV (ultraviolet) profiling data of 160 radix astragali samples from four geographic regions are trained and evaluated by four classification models, i.e., NNRW, MLP (multi-layer perceptron), SVM (support vector machine), and DTC (decision tree classifier). Their validation accuracies are 96.3%, 98.0%, 98.4%, and 92.8% respectively. The training/fitting times are 0.372 ms (milli-seconds), 57.9 ms, 2.033 ms, and 3.351 ms, respectively. This study shows that NNRW has a significant training time cut while keeping a high prediction accuracy, and it is a promising solution to resource-limited edge computing applications.

Variational Mode Decomposition Weighted Multiscale Support Vector Regression for Spectral Determination of Rapeseed Oil and Rhizoma Alpiniae Offcinarum Adulterants

The accurate prediction of the model is essential for food and herb analysis. In order to exploit the abundance of information embedded in the frequency and time domains, a weighted multiscale support vector regression (SVR) method based on variational mode decomposition (VMD), namely VMD-WMSVR, was proposed for the ultraviolet-visible (UV-Vis) spectral determination of rapeseed oil adulterants and near-infrared (NIR) spectral quantification of rhizoma alpiniae offcinarum adulterants. In this method, each spectrum is decomposed into K discrete mode components by VMD first. The mode matrix Uk is recombined from the decomposed components, and then, the SVR is used to build sub-models between each Uk and target value. The final prediction is obtained by integrating the predictions of the sub-models by weighted average. The performance of the proposed method was tested with two spectral datasets of adulterated vegetable oils and herbs. Compared with the results from partial least squares (PLS) and SVR, VMD-WMSVR shows potential in model accuracy.

Accurate identification of kidney injury progression via a fluorescent biosensor array

Identifying the progress of kidney injury may aid the effective treatment and intervention. Herein, we developed a fluorescent biosensor array for instantaneous and accurate identification of the kidney injury progression via "doubled" signals. The multichannel biosensor array consisted of polydopamine-polyethyleneimine (PDA-PEI) and multicolor-labelled different length of DNAs including AAAAA-Cyanine7 (5A-Cy7), AAAAAAAAAA-Texas Red (10A-Texas Red), and AAAAAAAAAAAAAAAAAAAA-VIC (20A-VIC). Facing to the variety of protein in urine with alterable charge accompanied with different progress of kidney injury, the composition of urine replaces the DNA signal molecules, forming their special fluorescence patterns. Taking the size of protein into consideration, the original three variables induced by the protein charge were extended to six variables induced by the two factors of protein particle size and charge difference, which could provide a more accurate strategy to identify the progress of kidney injury. Notably, this strategy not only opened up new perspective for identification the progress of kidney injury via the size and charge of urine protein, but also improved the resolving power of sensor array by increasing the number of sensor elements for extending their potential application to various diseases.

Application of ultraviolet-visible spectroscopy coupled with support vector regression for the quantitative detection of thiamethoxam in tea

A model combining UV-visible (UV-Vis) spectroscopy and support vector regression (SVR) for the quantitative detection of thiamethoxam in tea is proposed. First, each original UV-Vis spectrum in the sample set is decomposed into some intrinsic mode functions (IMFs) and a residual via ensemble empirical mode decomposition. Next, the decomposed IMFs are reconstructed into high-frequency and low-frequency matrices, and the residuals are combined into a trend matrix. Then, the SVR is used to build regression sub-models between each matrix and the content of thiamethoxam in tea. Finally, the combination model is established by a weighted average of the sub-models. The prediction results are compared with SVR and SVR coupled with several preprocessing methods, and the results demonstrate the superiority of the proposed approach in the quantitative detection of thiamethoxam in tea.

Optimization of sweet basil harvest time and cultivar characterization using near-infrared spectroscopy, liquid and gas chromatography, and chemometric statistical methods

BACKGROUND

Terpene, eugenol and polyphenolic contents of basil are major determinants of quality, which is affected by genetics, weather, growing practices, pests and diseases. Here, we aimed to develop a simple predictive analytical method for determining the polyphenol, eugenol and terpene content of the leaves of major Israeli sweet basil cultivars grown hydroponically, as a function of harvest time, through the use of near-infrared (NIR) spectroscopy, liquid/gas chromatography, and chemometric methods. We also wanted to identify the harvest time associated with the highest terpene, eugenol and polyphenol content.

RESULTS

Six different cultivars and four different harvest times were analyzed. Partial least square regression (PLS-R) analysis yielded an accurate, predictive model that explained more than 93% of the population variance for all of the analyzed compounds. The model yielded good/excellent prediction (R²  > 0.90, R² _cv and R² _pre  > 0.80) and very good residual predictive deviation (RPD > 2) for all of the analyzed compounds. Concentrations of rosmarinic acid, eugenol and terpenes increased steadily over the first 3 weeks, peaking in the fourth week in most of the cultivars. Our PLS-discriminant analysis (PLS-DA) model provided accurate harvest classification and prediction as compared to cultivar classification. The sensitivity, specificity and accuracy of harvest classification were larger than 0.82 for all harvest time points, whereas the cultivar classification, resulted in sensitivity values lower than 0.8 in three cultivars.

CONCLUSION

The PLS-R model provided good predictions of rosmarinic acid, eugenol and terpene content. Our NIR coupled with a PLS-DA demonstrated reasonable solution for harvest and cultivar classification. © 2021 Society of Chemical Industry.

Coal identification based on a deep network and reflectance spectroscopy

The rapid identification of coal types in the field is an important task. This research combines spectroscopy with deep learning algorithms and proposes a method for quickly identifying coal types in the field. First, we collect field spectral data of various coals and preprocess the spectra. Then, a coal identification model that uses a convolutional neural network in combination with an extreme learning machine is proposed. The two-dimensional spectral features of coal are extracted through the convolutional neural network, and the extreme learning machine is used as a classifier to identify the features. To further improve the identification performance of the model, we use the whale optimization algorithm to optimize the parameters of the model. The experimental results show that the proposed method can quickly and accurately identify types of coal. It provides a low-cost, convenient, and effective method for the rapid identification of coal in the field.

Multi-platform integration based on NIR and UV-Vis spectroscopies for the geographical traceability of the fruits of Amomum tsao-ko

Due to the world-wide concern relating to herb quality and safety, there is a momentum to authenticate the geographical origin of herb with multi-platform techniques. This study attempted to assess multi-platform information as a practical strategy for the geographical traceability of the fruits of Amomum tsao-ko. To this aim, one hundred and eighty dried fruits of A. tsao-ko from five geographical regions were analyzed by near infrared (NIR) and ultraviolet visible (UV-Vis) spectroscopy. On this basis, two variable dimension reduction strategies, including principal component analysis (PCA) and sequential and orthogonalized partial-least squares (SO-PLS), and two variables selection strategies, including variable importance in projection (VIP) and sequential and orthogonalized covariance selection (SO-CovSel), were performed to extract the feature information in the two blocks. Partial least squares discriminant analysis (PLS-DA) classification algorithm combined with fused matrices was used to identify the geographical origins. The results of PLS-DA models indicated that SO-PLS and SO-CovSel, taking advantage of the sequential modeling coupled to orthogonalization, could not only identify the common information presented in the two blocks but also provide more concise methods without any loss of classification ability, which could be employed in authenticating the geographical regions of the fruits of A. tsao-ko, effectively.

Fast discrimination and quantification analysis of Curcumae Radix from four botanical origins using NIR spectroscopy coupled with chemometrics tools

Curcumae Radix (Yujin) is a multi-origin herbal medicine with excellent clinical efficacy. For fast discrimination and quantification analysis of Yujin from four botanical origins (Guiyujin, Huangyujin, Lvyujin and Wenyujin), near infrared (NIR) spectroscopy combined with chemometrics tools was employed in this study. Based on NIR data, principal component analysis (PCA) could only realize the separation between Guiyujin and Wenyujin samples, and the partial least squares-discrimination analysis (PLS-DA), support vector machine (SVM) and k-nearest neighbors (KNN) models achieved the complete discrimination of the four species of Yujin with 100% accuracy. Moreover, the method for the simultaneous determination of six bioactive compounds in Yujin was developed by HPLC. Germacrone, curdione and curcumenol could be found in all samples, and curcumin, demethoxycurcumin and bisdemethoxycurcumin were only observed in Huangyujin samples. Then, the support vector machine regression (SVMR) model for the prediction of germacrone content was successfully constructed. And the coefficients of determination were 0.88 and 0.89 for calibration and validation sets, respectively. The present work proposes a quick, economic and reliable method for the discrimination of Yujin from four botanical origins and the prediction of germacrone content, which will contribute to its quality control researches.

UV spectroscopy and HPLC combined with chemometrics for rapid discrimination and quantification of Curcumae Rhizoma from three botanical origins

Curcumae Rhizoma (Ezhu in Chinese) is a multi-origin herbal medicine with excellent clinical efficacy. For fast discrimination and quantification analysis of Ezhu from three botanical origins (Curcuma kwangsiensis, Curcuma phaeocaulis, and Curcuma wenyujin), ultra-violet (UV) spectroscopy and high performance liquid chromatography (HPLC) combined with chemometric tools were employed in this study. Firstly, the analysis method for the simultaneous determination of eleven compounds in Ezhu was developed by HPLC, and the UV spectra of thirty-eight batches of Ezhu were acquired. Then, principal component analysis (PCA), an unsupervised pattern recognition method, was applied on the HPLC and UV spectral data. PCA did not show a clear separation between C. phaeocaulis and C. wenyujin samples with HPLC data. By contrast, the supervised techniques, decision tree (DT) and linear discriminant analysis (LDA), achieved the complete discrimination for the three species of Ezhu with 100 % correct classification rate (CCR), showing excellent performance. Based on UV spectral data, PCA presented good performance for discriminating the three species of Ezhu. LDA, support vector machine (SVM) and k-nearest neighbors (KNN) models provided 96.3 % CCR for the calibration set and 100 % CCR for the validation set. Moreover, the partial least squares (PLS) and back propagation-artificial neural network (BP-ANN) quantitative models established on UV spectral data were satisfactory in predicting the contents of zederone, curdione and 3,5-dihydroxy-1-(3,4-dihydroxyphenyl)-7-(4-hydroxyphenyl)-heptane. The residual predictive deviation (RPD) for zederone, curdione and 3,5-dihydroxy-1-(3,4-dihydroxyphenyl)-7-(4-hydroxyphenyl)-heptane of PLS models were 3.169, 1.502 and 1.735, and that of BP-ANN models were 3.467, 2.481 and 2.370, respectively. The present work proposed a rapid and reliable method for the discrimination of Ezhu from three botanical origins and the prediction of zederone, curdione and 3,5-dihydroxy-1-(3,4-dihydroxyphenyl)-7-(4-hydroxyphenyl)-heptane contents in Ezhu, which will help a lot in the quality control of Ezhu and other multi-origin herbal medicines.

A nanosensor for precise discrimination of nephrotoxic drug mechanisms via dynamic fluorescence fingerprint strategy

Drug-induced kidney injury causes structural or functional abnormalities of kidney, seriously affecting clinical practice and drug discovery. However, rapid and effective identification of nephrotoxic drug mechanisms is yet a challenging task arising from the complexity and diversity of various nephrotoxic mechanisms. Herein, we have constructed a polydopamine-polyethyleneimine/quantum dots sensor to instantaneously read out the nephrotoxic drugs mechanisms based on the disparate cell surface phenotypes. Cell surface components induced by multiple nephrotoxic drugs can change the fluorescence emission of multicolor quantum dots, generating their corresponding fluorescent fingerprints. The fluorescence response signatures induced by different nephrotoxic agents are gained with 84% accuracy via linear discriminant analysis. Furthermore, taking the time-toxicity relationship into consideration, dynamic fluorescent fingerprint is obtained through continuous monitoring the progress of renal cell damage, achieving 100% precise classification for nephrotoxic mechanisms of four types of antibiotics. Notably, the fluorescent fingerprint-based high-throughput sensor has been demonstrated by successfully distinguishing nephrotoxic drugs in seconds, employing a promising protocol to discriminate the specific mechanism of nephrotoxic drugs, as well as drug safety evaluation.

Weighted multiscale support vector regression for fast quantification of vegetable oils in edible blend oil by ultraviolet-visible spectroscopy

Weighted multiscale support vector regression combined with ultraviolet-visible (UV-Vis) spectra for quantitative analysis of edible blend oil is proposed. In the approach, UV-Vis spectra of the training set are decomposed into a certain number of intrinsic mode functions (IMFs) and a residue by empirical mode decomposition (EMD) at first. Then support vector regression (SVR) sub-models are built on each IMF and residue. For prediction set, the spectra are decomposed as done on the training set and the final predictions are obtained by integrating SVR sub-model predictions by weighted average. The weight of the sub-model is the reciprocal of the fourth power of the root mean square error of cross-validation (RMSECV). For predicting peanut oil in binary blend oil and sesame oil in ternary blend oil, the proposed method has superiority in root mean square error of prediction (RMSEP) and correlation coefficient (R) compared with SVR and partial least squares (PLS).