Multivariate Partial Least Squares Regression Research Articles

High-Precision and Rapid In Situ Ore Element Detection Based on Laser-Induced Breakdown Spectroscopy

In most cases, in situ online composition detection is highly desirable for the iron and steel industry. High-precision prediction of the iron content of an iron ore is quite difficult due to the complexity of ore composition. Herein, an online composition analysis system based on laser-induced breakdown spectroscopy (LIBS) is built up for real-time Fe concentration determination. Subsequently, using the support vector machine (SVM) combined with multivariate partial least-squares regression (PLSR) to establish a linear relationship between the spectral data and typical element content, the iron content in the standard ore tablet is used as the true value to train the model, and the iron content in the raw ore block is predicted, with an error of 1.6%. By studying the elemental distribution content of the raw ore with repeated laser ablation, it is found that the internal element content of the ore changes at different depths from the surface, and the element content quickly stabilizes. The results demonstrate that the method can accurately and effectively predict iron content online, allowing the application of online detection of industrial ore composition.

Improving forest burn severity estimations with partial least squares regression and orthogonal signal correction methods in Daxing’an Mountains, China

Several indices and simple empirical models and ratios of single band from pre- and post-fire Landsat images have been developed to estimate and/or map burn severity. However, these models and indices are usually site-, time- and vegetation-dependent and their applications are limited. The Daxing’an Mountains range has the largest forested area in China and is prone to wildfires. Whether or not the existing models can effectively characterize the burn severity over a large region is unclear. In this study, we used the orthogonal signal correction method based on partial least squares regression (PLSR) to select those variables that better interpret the variance of burn severity. A new index and other commonly used indices were used to construct a new, multivariate PLSR model which was compared with the popular single variable models, according to three assessment indices: relative root mean square error (RMSE%), relative bias (RE%) and Nash–Sutcliffe efficiency (NSE%). The results indicate that the multivariate PLSR model performed better than the other single variable models with higher NSE% (68.2% vs. 67.8%) and less RE% (3.7% vs. − 8.7%), while achieving almost the same RMSE%. We also discuss the spectral characteristics of the four selected variables for constructing the multivariate PLSR model and their correlation with the field burn severity data. The new model developed from this study should help to better understand the patterns of forest burn severity and assist in vegetation restoration efforts in the region.

Modeling the soil burn effect for temperature prediction by energy dispersive X ray Fluorescence in an haplic cambisol soil

Considered an important hydrological and geomorphological agent, fire can cause physical, chemical and biological changes in the soil. Besides wildfire, the study of fire effects is also related to traditional agriculture. This study presents the characterization and analysis of soil samples submitted to burn simulation with the objective to build a temperature prediction model in order to determine the maximum temperature reached by real soil burn samples. For this purpose, surface soil samples (0–2.5 cm) classified as Haplic Cambisol were collected from a native forest area close to the studied field. The temperature of the simulation samples ranged from 50 to 750 °C. Moreover, a real burn set of samples were measured for temperature prediction using the proposed model. The characterization and quantification of the chemical elements present in the soil were done by Energy Dispersive X-ray Fluorescence (EDXRF) measurements. Plots with Fe concentration and with the Rayleigh and Compton scattering data versus temperature were constructed. The Fe/Rh and Fe/RhC ratios resulted in relative deviations ranging from 14% to 22% using univariate analysis. Multivariate analysis was also applied through partial least squares regression (PLSR) method in four different spectrum regions. The best result was obtained for the model using the spectrum scattering region with r2 = 0.90 and relative deviation ranging from 8% to 25% for the predicted temperature. The use of local multivariate PLSR models improved the results when compared to the univariate regression results.

Feasibility of attenuated total reflection-fourier transform infrared (ATR-FTIR) chemical imaging and partial least squares regression (PLSR) to predict protein adhesion on polymeric surfaces.

Predicting the degree to which proteins adhere to a polymeric surface is an ongoing challenge in the scientific community to prevent non-specific protein adhesion and drive favourable protein - surface interactions. This work explores the potential of multivariate PLSR modelling in conjunction with Attenuated Total Reflection - Fourier Transform Infrared (ATR-FTIR) chemical imaging to investigate whether experimentally characterised surface chemistry can be used to predict surface protein adhesion. ATR-FTIR spectra were collected on dry and wetted polymeric surfaces, followed by evaluation of adhered fibrinogen on surfaces using the micro bicinchoninic (BCA) protein assay as a reference method. Partial Least Squares Regression (PLSR) models were built using IR spectra as the predictor variable. Overall the models built with 'wetted polymer' IR spectra performed better as compared to the models built using 'dry polymer' IR spectra (average coefficient of determination, R2P 0.998, 0.996 respectively), with the lowest error in prediction (4 ± 0.6 μg) for ultra-high molecular weight polyethylene (UHMPE) as a test surface. This indicates the potential of this method to predict the degree to which protein adhesion occurs on polymeric surfaces using experimentally determined surface chemistry.

Discriminating growth stages of an endangered Mediterranean relict plant (Ammopiptanthus mongolicus) in the arid Northwest China using hyperspectral measurements

Ammopiptanthus mongolicus, the only drought-resistant, leguminous, evergreen shrub in the desert region of China, is endangered due to climate change and its growth stages urgently need to be non-destructively detected. Although many spectral indexes have been proposed for characterizing vegetation, the relationships are often inconsistent, making it challenging to characterize the status of vegetation across all growth stages. This study investigated the Spectral Features of the endangered desert plant A. mongolicus at different growth stages, and extracted the identified Spectral Features for the establishment of detection and discrimination models using Partial Least Square Regression (PLSR) and Fisher Linear Discriminate Analysis (FLDA), respectively. The results showed spectral reflectance of A. mongolicus differed across different growth stages and it generally increased with the degree of senescence. Poor performance was found in the single factor model, with RMSE ranging from 20.34 to 27.39 or Overall Accuracy of 60% in the validation datasets. The multivariate PLSR model, based on Normalized Difference Vegetation Index (NDVI), Normalized Difference Red Edge Index (NDRE), Physiological Reflectance Index (PhRI) and Plant Senescence Reflectance Index (PSRI), turned out to be accurate in detecting the growth stages, with R2 of 0.89 and RMSE of 12.46, and the performance of the multivariate FLDA model based on 14 Spectral Features was acceptable, with an Overall Accuracy of 89% in the validation datasets. This research provides useful insights for timely and non-destructively discriminating different growth stages by using multivariate PLSR and FLDA analysis.

Online low-field NMR spectroscopy for process control of an industrial lithiation reaction-automated data analysis.

Monitoring specific chemical properties is the key to chemical process control. Today, mainly optical online methods are applied, which require time- and cost-intensive calibration effort. NMR spectroscopy, with its advantage being a direct comparison method without need for calibration, has a high potential forenabling closed-loop process control while exhibiting short set-up times. Compact NMR instruments make NMR spectroscopy accessible in industrial and rough environments for process monitoring and advanced process control strategies. We present a fully automated data analysis approach which is completely based on physically motivated spectral models as first principles information (indirect hard modeling-IHM) and applied it to a given pharmaceutical lithiation reaction in the framework of the European Union's Horizon 2020 project CONSENS. Online low-field NMR (LF NMR) data was analyzed by IHM with low calibration effort, compared to a multivariate PLS-R (partial least squares regression) approach, and both validated using online high-field NMR (HF NMR) spectroscopy. Graphical abstract NMR sensor module for monitoring of the aromatic coupling of 1-fluoro-2-nitrobenzene (FNB) with aniline to 2-nitrodiphenylamine (NDPA) using lithium-bis(trimethylsilyl) amide (Li-HMDS) in continuous operation. Online 43.5 MHz low-field NMR (LF) was compared to 500 MHz high-field NMR spectroscopy (HF) as reference method.

Dynamic 2-Deoxy-2-[18F]Fluoro-D-Glucose Positron Emission Tomography for Chemotherapy Response Monitoring of Breast Cancer Xenografts.

Non-invasive response monitoring can potentially be used to guide therapy selection for breast cancer patients. We employed dynamic 2-deoxy-2-[18F]fluoro-D-glucose positron emission tomography ([18F]FDG PET) to evaluate changes in three breast cancer xenograft lines in mice following three chemotherapy regimens. Sixty-six athymic nude mice bearing bilateral breast cancer xenografts (two basal-like and one luminal-like subtype) underwent three 60min [18F]FDG PET scans. Scans were performed prior to and 3 and 10days after treatment with doxorubicin, paclitaxel, or carboplatin. Tumor growth was monitored in parallel. A pharmacokinetic compartmental model was fitted to the tumor uptake curves, providing estimates of transfer rates between the vascular, non-metabolized, and metabolized compartments. Early and late standardized uptake values (SUVE and SUVL, respectively); the rate constants k 1, k 2, and k 3, and the intravascular fraction v B were estimated. Changes in tumor volume were used as a response measure. Multivariate partial least-squares regression (PLSR) was used to assess if PET parameters could model tumor response and to identify PET parameters with the largest impact on response. Treatment responders had significantly larger perfusion-related parameters (k 1 and k 2) and lower metabolism-related parameter (k 3) than non-responders 10days after the start of treatment. These findings were further supported by the PLSR analysis, which showed that k 1 and k 2 at day 10 and changes in k 3 explained most of the variability in response to therapy, whereas SUVL and particularly SUVE were of lesser importance. Overall, rate parameters related to both tumor perfusion and metabolism were associated with tumor response. Conventional metrics of [18F]FDG uptake such as SUVE and SUVL apparently had little relation to tumor response, thus necessitating full dynamic scanning and pharmacokinetic analysis for optimal evaluation of chemotherapy-induced changes in breast cancers.

Analysis of slags using laser-induced breakdown spectroscopy

The feasibility of laser-induced breakdown spectroscopy (LIBS) for the analysis of gasification slags was investigated by comparing LIBS results to the results of an ICP-OES analyzer. A small amount of slag sample was placed on a piece of double sided adhesive tape attached to a glass microscope slide and analyzed for Al, Ca, Fe, Si, and V which are major elements found in slags. The partial least squares regression (PLS-R) and univariate simple linear regression (SLR) calibration methods indicated that apart from V (accuracy up to +20%) the accuracy of analysis varies within 0.35–6.5% for SLR and 0.06–10% for PLS-R. A paired-sample t-test within the 95% confidence level yielded p-values greater than 0.05, meaning no appreciable statistical difference was observed between the univariate SLR with internal standardization and the multivariate PLS-R for most of the analytes. From the results obtained in this work, LIBS response varies depending on the element and the technique used for quantitative analysis. Simultaneous use of the univariate calibration curves with internal standard (intensity ratio) and PLS regression in multi-elemental analysis can help reduce the matrix effect of slags associated to their high variation in concentration. Overall, these results demonstrate the capability of LIBS as an alternative technique for analyzing gasification slags. Estimated limits of detection for Al, Ca, Fe, Si and V were 0.167, 0.78, 0.171, 0.243 and 0.01wt.%, respectively.

Laser ablation molecular isotopic spectrometry of water for 1D2/1H1 ratio analysis

Laser Ablation Molecular Isotopic Spectrometry (LAMIS) has been investigated for optical isotopic analysis of the deuterium to protium ratio in enriched water samples in ambient air at atmospheric pressure. Multivariate PLSR (Partial Least Squares Regression) based calibrations were carried out and validated using multiple statistical parameters. Comparisons of results are reported using two spectrometers having two orders of magnitude difference in spectral resolution. The accuracy and precision of isotopic analysis depends on the spectral resolution and the inherent isotope shift of the elements. The requirements for spectral resolution of the measurement system can be significantly relaxed when the isotopic abundance ratio is determined using chemometric processing of the spectra. Large isotopic shifts in the individual rotational branches of OH/OD molecular emission spectra were measured. Optimized temporal conditions for LAMIS measurements were established. Several sub-regions of spectra were used for PLSR calibration and the results demonstrate that both the emission intensity and degree of spectral differentiation affect the quality of the PLSR calibration. LAMIS results also were compared with traditional LIBS results obtained using PLSR and a spectral deconvolution method, demonstrating the advantages of LAMIS over LIBS with respect to isotopic composition determination.

Predictive factors for disability outcome at twenty weeks and two years following a pain self-management group intervention in patients with persistent neck pain in primary health care

To explore possible predictors associated with short-term (post-treatment) and long-term (2 years) treatment success in terms of pain-related disability for patients with persistent neck pain following a pain and stress self-management intervention (PASS). Data from 77 participants assigned to PASS in a randomized controlled trial were explored to identify possible predictors of favourable outcome regarding pain-related disability as measured by the Neck Disability Index (NDI), by use of Pearson correlation analysis, partial least squares (PLS) and ordinary least squares (OLS) regression analyses. Data from self-assessment questionnaires completed by the participants before, post-treatment (i.e. 20 weeks after inclusion) and 2 years after inclusion in the study, were used. Multivariate PLS regression analysis showed that baseline scores in NDI, the Self-Efficacy Scale (SES) and pain intensity explained 31% of the variance in disability (NDI) post-treatment. Multivariate PLS regression analysis showed that post-treatment scores in NDI, SES and pain intensity explained 68% of the variance in disability (NDI) at 2 years. Treatment gains, as measured by post-treatment scores at 20-week follow-up, in disability, self-efficacy and pain intensity were associated with long-term outcome in pain-related disability at 2 years, in patients with persistent neck pain participating in a self-management group intervention in primary health care.

Performance of PLS regression coefficients in selecting variables for each response of a multivariate PLS for omics-type data

Multivariate partial least square (PLS) regression allows the modeling of complex biological events, by considering different factors at the same time. It is unaffected by data collinearity, representing a valuable method for modeling high-dimensional biological data (as derived from genomics, proteomics and peptidomics). In presence of multiple responses, it is of particular interest how to appropriately "dissect" the model, to reveal the importance of single attributes with regard to individual responses (for example, variable selection). In this paper, performances of multivariate PLS regression coefficients, in selecting relevant predictors for different responses in omics-type of data, were investigated by means of a receiver operating characteristic (ROC) analysis. For this purpose, simulated data, mimicking the covariance structures of microarray and liquid chromatography mass spectrometric data, were used to generate matrices of predictors and responses. The relevant predictors were set a priori. The influences of noise, the source of data with different covariance structure and the size of relevant predictors were investigated. Results demonstrate the applicability of PLS regression coefficients in selecting variables for each response of a multivariate PLS, in omics-type of data. Comparisons with other feature selection methods, such as variable importance in the projection scores, principal component regression, and least absolute shrinkage and selection operator regression were also provided.

Process analytical technology: Non-destructive assessment of anastrozole entrapment within PLGA microparticles by near infrared spectroscopy and chemical imaging

The objective of this study was to evaluate near-infrared (NIR) spectroscopy and imaging as approaches to assess anastrozole entrapment within PLGA microparticles. By varying the polymer concentration, three batches containing the same amount of the drug were prepared. The spectral features that allow NIR drug quantitation were evaluated and compared with a best fit line algorithm. Actual entrapment efficiencies (EEF) determined via a destructive method were used for construction of calibration models using partial least square regression (PLS) or the algorithm. On the other hand, a chemical imaging system based on array detector technology was used to rapidly collect high contrast NIR images of the formulated microparticles. A quantitative measure of anastrozole entrapped was determined by calculating the percentage standard deviation of the distribution of pixel intensities in the PLS score images and histograms. Concerning conventional NIR analysis, both methods were equivalent for the prediction of EEF over the range of polymer levels studied. Correlation coefficients of more than 0.992 were obtained for either the calibration or prediction of EEF by the two methods; 0.392% and 0.374% were the standard errors of calibration and prediction (SEC and SEP) obtained for the prediction of EEF using the fit line, respectively, whereas the prediction of the EEF by the partial least square regression showed a SEC of 0.195% and SEP of 0.179%. As a result, the spectral best fit algorithm method compared favourably to the multivariate PLS method, but was easier to develop. In contrast, NIR spectral imaging was capable of clearly differentiating the three batches, both qualitatively and quantitatively. The percentage standard deviation increased progressively by increasing the ratio of drug-to-polymer concentrations. In conclusion, both NIR approaches were capable of accurate assessment of drug entrapment within microparticles. In addition, the NIR spectral imaging system provides a rapid approach for acquiring spatial and spectral information on microparticles.

Determination of tacticity in polypropylene by FTIR with multivariate calibration

AbstractA method for determination of tacticity in polypropylene (PP) using FTIR associated with multivariate analysis is presented. Blends of PP with known tacticity were prepared with isotactic, syndiotactic, and atactic polymer and analyzed by 13C‐NMR. The FTIR spectra were recorded and processed through principal components regression (PCR) and partial least‐squares regression (PLS), using information from several different portions of the spectra. The method was compared with the classical methods of tacticity determination by FTIR based on the intensities of the bands at 998 cm−1 (isotactic), 868 cm−1 (syndiotactic), and 975 cm−1 (internal standard), which are known to be dependent on the crystallinity of the polymer and, thus, affected by temperature and sample preparation. The models obtained with multivariate calibration, both with PCR and PLS, gave prediction errors up to fivefold smaller than that of the classical methods, and were also shown not to be heavily dependent on the bands that are affected by the crystallinity of the polymer, but rather on the methyl and methylene bendings at 1375 and 1462 cm−1. © 2002 Wiley Periodicals, Inc. J Appl Polym Sci 85: 734–745, 2002