Selection Of Spectral Regions Research Articles

Improving Raman-Based Models for Real-Time Monitoring the CHO Cell Culture Process with Effective Variable Selection Strategies

Research has shown that Raman spectroscopy could be applied to monitor various components in mammalian cell culture in real time. In the process of application, it is necessary to ensure the performance of the Raman-based model. The variable selection strategy is an effective method that significantly influences the model performance and simplification. In this study, different variable selection strategies were evaluated, and the optimal variable selection strategy was determined for monitoring the CHO cell culture process. Firstly, a wide variety of spectral regions involving the Raman fingerprinting region and the C-H stretching region were investigated. Secondly, six different variable selection algorithms were meticulously assessed. Thirdly, the combination of different variable selection algorithms was used to improve model performance and simplify the model. Finally, the monitoring of cell culture processes was implemented. The findings underscored that commonly used spectral regions could improve the model performance but could not simplify the model well. Moving-window partial least square (MWPLS), genetic algorithm (GA), and random frog (RF) are more suitable for Raman modeling of the cell culture process, but they must be used after the spectral region selection. The combination of three variable selection algorithms (MWPLS-GA-RF) improved the model’s performance by 16–70% by selecting 30–60 variables, effectively simplifying the model. For glucose, lactate, viable cell density, and ammonium ion, real-time monitoring was performed well. This study will be helpful for researchers to select suitable variable selection strategies for building models for the real-time monitoring of cell culture.

Multivariate calibration strategy in simultaneous determination of temperature properties of petroleum diesel by near infrared spectrometry

It is essential to control the quality of diesel products so that they comply with relevant fuel specifications, however, the quality assessments rely upon conventional wet chemical analyses that are costly and time consuming. Rapid, simultaneous quality measurement enabling immediate online optimisation for process control and blending offers tremendous cost savings by minimising product quality give-away, shipment demurrage, tank inventory, and laboratory analysis. In this study, the use of near infrared spectroscopy and chemometrics demonstrates a straightforward workflow for simultaneous determination of the petroleum diesel’s boiling point at 95% recovery (T95), flash point (FP), cloud point (CP), and cetane index (CI) calibration development. It involved appropriate spectral region selection, calibration/validation set partition, data pre-processing, regression modelling and validation. Based on the calibration and validation results, the supervised learning models that are obtained from a combination region of 4000–4800 cm−1 on a randomly selected calibration set managed to deliver promising predictive performance in terms of coefficient of determination for prediction (r2P/T95 ≥ 0.94, r2P/FP ≥ 0.89, r2P/CP ≥ 0.89, r2P/CI ≥ 0.993), root mean square error of prediction (RMSEP (T95) ≤ 5.2°C, RMSEP (FP) ≤ 2.0°C, RMSEP (CP) ≤ 2.4°C, RMSEP (CI) ≤ 0.3), and ratio of performance deviation (RPD (T95) ≥ 3.7, RPD (FP) ≥ 3.0, RPD (CP) ≥ 2.9, RPD (CI) ≥ 11). Regardless of principal component regression or partial least square regression on either the multiplicative scattering corrected spectra or Savitzky-Golay second derivative spectra, the developed models met respective ASTM reproducibility requirements, and were considered adequate for immediate quality assessment of diesel.

Influence of polarized reflection on airborne remote sensing of canopy foliar nitrogen content

ABSTRACT Radiation detected by sensors can reveal information regarding canopy structure and leaf surface properties. It has been shown that the scattering coefficient (Wλ ), a ‘pure’ spectrum obtained after correction for the effect of canopy structure on the bidirectional reflectance factor (BRF), can provide information about biochemical composition; e.g. nitrogen content, more objectively. However, specular radiation reflected from leaf surface (without any information from the leaf interior) has a potential influence on Wλ and thus consequently affects the estimation of canopy nitrogen content (CNC). The so-called specular reflection influence has rarely been studied. Polarized reflectance (R p), as part of the specular reflectance, was used to assess this kind of influence. To model the accurate R p within the study area, six bidirectional polarization distribution function (BPDF) models were intercompared, using the existing POLDER/PARASOL R p database, to produce R p for three vegetation types: needleleaf, broadleaf, and mixed forest. Then, correction for polarized reflection and canopy structure was made to yield a more reasonable and accurate Wλ . An efficient interval partial least-squares regression (iPLSR) method for CNC estimation was used to analyse the impact of polarized reflection on both the 400 nm–2,500-nm canopy Wλ and CNC estimation. The results showed that the contribution of polarization was greater in the strong-absorption spectral region, with an influence on Wλ estimation of up to 25% and 28% in the visible and short-wave infrared regions, respectively. Also, improvements of 0.93% in average regarding the relative mean square error of cross-validation (RMSECV) were seen in CNC estimation accuracy. Moreover, sensitivity analyses of BPDF model parameters, and of the fixed total number of intervals in the iPLSR were also conducted. Improvements with an average of 1.19% regarding the RMSECV was consistently seen in the accuracy of CNC regression after polarization correction, no matter how many fixed intervals were. The results also gave guides on the selection of spectral regions used to estimate forest CNC.

Excitation energy transfer kinetics and efficiency in phototrophic green sulfur bacteria

A series of spectroscopic measurements were performed on membrane fractions and detergent-solubilized complexes from the green sulfur bacterium (GSB) Chlorobaculum (Cba.) tepidum. The excitation migration through the entire GSB photosynthetic apparatus cannot be observed upon excitation of membranes in the chlorosome region at 77 K. In order to observe energy transfer from the Fenna-Matthews-Olson (FMO) protein to the reaction center (RC), FMO was directly excited at ~800 nm in transient absorption experiments. However, interpretation of the results is complicated by the spectral overlap between FMO and the RC. The availability of the Y16F FMO mutant, whose absorption spectrum is drastically different from that of the WT, has enabled the selection of spectral regions where either only FMO or the RC contributes. The application of a directed kinetic modeling approach, or target analysis, revealed the various decay and energy transfer pathways within the pigment-protein complexes. The calculated FMO-to-RC excitation energy transfer efficiencies are approximately 25% and 48% for the Y16F and WT samples, respectively.

Quantification of Soil Permanganate Oxidizable C (POXC) Using Infrared Spectroscopy

Core Ideas Infrared spectroscopy and chemometrics can be used to quantify soil labile C. Soils amended with organic materials need to be taken into account during calibration development. Selection of spectral regions can give parsimonious predictions of soil total and labile C. Labile soil carbon is an important component of soil organic matter because it embodies the mineralizable material that is associated with short‐term fertility. Permanganate‐oxidizable C (POXC) is a widely used method for the study of labile C dynamics in soils. Rapid methods are needed to measure labile C, and better understand how this pool varies with soil C at regional scales. Infrared spectroscopy is an inexpensive way to quantify SOC and observe fluctuations in C functional groups. Using a sample set that encompassed several soil types and plant communities (seven different research projects, n = 496), soils were analyzed via diffuse reflectance Fourier transformed mid‐infrared (MidIR, 4000–400 cm–1) and near‐infrared (NIR, 10000–4000 cm–1) spectroscopy. Spectral data were used to develop calibrations for POXC, soil organic C (SOC), and total N (TN) using partial least squares (PLS) regression. The MidIR predicted POXC slightly better than the NIR, with calibration and/or validation R2 values ranging from 0.77 to 0.81 depending on spectral pretreatments. Predictions for POXC were better than SOC and TN, but site variability influenced the calibration quality for SOC and TN. Using a selected MidIR region, which included bands correlated to POXC (3225–2270 cm–1), reduced the calibration quality, but still gave acceptable R2 values of 0.76 to 0.77 for the calibration and validation sets. We show that POXC can be predicted using NIR and MidIR spectra. Selecting informative spectral bands offers an alternative to using full spectra for PLS regressions.

Application of fluorescence spectroscopy combined with interval partial least squares to the determination of enantiomeric composition of tryptophan

The application of interval partial least squares (iPLS) method to the fluorescence spectroscopy analysis was investigated. The practicability of spectral region selection in the fluorescence spectroscopy analysis was studied. A method which combines fluorescence spectroscopy and iPLS was proposed for determining the enantiomeric composition of tryptophan (Trp). Fluorescence spectroscopy was used to measure the diastereomeric interaction between the enantiomers of Trp and bovine serum albumin, which plays the role of chiral selector. iPLS was used to select the spectral region and build the calibration model between the fluorescence spectral data and enantiomeric composition of Trp. The spectral region from 300nm to 356nm was selected and used to build the calibration model. Leave-one-out cross validation and external test validation were used to validate the obtained models. By selecting the spectral region, iPLS provided better prediction performance, in comparison to the full-spectrum PLS model. The root mean square relative error (RMSRE) of leave-one-out validation has decreased from 10.24 to 7.17, and the RMSRE of external test validation has decreased from 8.59 to 7.06. In addition, four local-spectrum PLS models were developed in order to make a comparison to the iPLS model. The result demonstrates the iPLS model is superior to the four local-spectrum PLS models. The result of this work demonstrates the proposed method is practicable for determining the enantiomeric composition of Trp at trace level. When there is 2.50μmol·L−1 Trp in the samples, the enantiomeric composition can be accurately determined. Moreover, the result demonstrates the selection of spectral region has significant influence on the fluorescence spectroscopy analysis and iPLS is a practicable method for spectral region selection.

Green method by diffuse reflectance infrared spectroscopy and spectral region selection for the quantification of sulphamethoxazole and trimethoprim in pharmaceutical formulations.

An alternative method for the quantification of sulphametoxazole (SMZ) and trimethoprim (TMP) using diffuse reflectance infrared Fourier-transform spectroscopy (DRIFTS) and partial least square regression (PLS) was developed. Interval Partial Least Square (iPLS) and Synergy Partial Least Square (siPLS) were applied to select a spectral range that provided the lowest prediction error in comparison to the full-spectrum model. Fifteen commercial tablet formulations and forty-nine synthetic samples were used. The ranges of concentration considered were 400 to 900 mg g-1SMZ and 80 to 240 mg g-1 TMP. Spectral data were recorded between 600 and 4000 cm-1 with a 4 cm-1 resolution by Diffuse Reflectance Infrared Fourier Transform Spectroscopy (DRIFTS). The proposed procedure was compared to high performance liquid chromatography (HPLC). The results obtained from the root mean square error of prediction (RMSEP), during the validation of the models for samples of sulphamethoxazole (SMZ) and trimethoprim (TMP) using siPLS, demonstrate that this approach is a valid technique for use in quantitative analysis of pharmaceutical formulations. The selected interval algorithm allowed building regression models with minor errors when compared to the full spectrum PLS model. A RMSEP of 13.03 mg g-1for SMZ and 4.88 mg g-1 for TMP was obtained after the selection the best spectral regions by siPLS.

NIR pre-selection data using modified changeable size moving window partial least squares and pure spectral chemometrical modeling with ant colony optimization for wheat flour characterization

The aim of process optimization is obtaining higher productivity and profit in chemical or bio-chemical process. For that, one must apply control techniques that closely correlate with our ability to characterize a process. Optical sensors associated with chemometric modeling are considered a natural choice for non-intrusive and high sensitivity measurements. This study focus on wheat flour characterization (usual and mandatory action, widely present on the food industry) using near-infrared, comparing two approaches for spectral region selection: modified CSMWPLS and PSCM/ACO. Spectroscopic data is assayed using a combination of CSMWPLS and variable selection algorithm based on ant colony optimization. Protein prediction results are compared with standards PLS, CSMWPLS and PSCM/ACO models. Prediction capability improved 46% using modified CSMWPLS and PSCM/ACO modeling, confirming the efficiency of the proposed characterization methods and chemometric modeling strategy.

Cross-scale predictive modeling of CHO cell culture growth and metabolites using Raman spectroscopy and multivariate analysis.

Multi-component, multi-scale Raman spectroscopy modeling results from a monoclonal antibody producing CHO cell culture process including data from two development scales (3 L, 200 L) and a clinical manufacturing scale environment (2,000 L) are presented. Multivariate analysis principles are a critical component to partial least squares (PLS) modeling but can quickly turn into an overly iterative process, thus a simplified protocol is proposed for addressing necessary steps including spectral preprocessing, spectral region selection, and outlier removal to create models exclusively from cell culture process data without the inclusion of spectral data from chemically defined nutrient solutions or targeted component spiking studies. An array of single-scale and combination-scale modeling iterations were generated to evaluate technology capabilities and model scalability. Analysis of prediction errors across models suggests that glucose, lactate, and osmolality are well modeled. Model strength was confirmed via predictive validation and by examining performance similarity across single-scale and combination-scale models. Additionally, accurate predictive models were attained in most cases for viable cell density and total cell density; however, these components exhibited some scale-dependencies that hindered model quality in cross-scale predictions where only development data was used in calibration. Glutamate and ammonium models were also able to achieve accurate predictions in most cases. However, there are differences in the absolute concentration ranges of these components across the datasets of individual bioreactor scales. Thus, glutamate and ammonium PLS models were forced to extrapolate in cases where models were derived from small scale data only but used in cross-scale applications predicting against manufacturing scale batches.

Enhanced Characterization of Naproxen Formulation by Near Infrared Spectroscopy

Near infrared spectroscopy in combination with appropriate chemometric methods is an effective technique for quantitative analysis of parameters of interest for the pharmaceutical industry. In this study, the artificial neural network (ANN) was applied to monitor critical parameters (compression force, tablet hardness, mean particle size, and active pharmaceutical ingredient concentration of tablets) in the process of naproxen pharmaceutical preparation. The performance of ANN was compared to linear methods (partial least squares regression (PLS) and synergy interval partial squares (siPLS)). The ANN models for compression force, tablet hardness, mean particle size, and active pharmaceutical ingredient concentration of tablets yielded the low root mean square error of prediction (RMSEP) values of 0.936 KN, 0.302 kg, 4.49 mg, and 2.14 µm, respectively. The predictive ability of the PLS model was improved by siPLS with selection of spectral regions and the best performance among all calibration methods was showed by the nonlinear method (ANN). Effective models were built by using these approaches using near infrared spectroscopy.

Simultaneous determination of acesulfame-K, aspartame and stevioside in sweetener blends by ultraviolet spectroscopy with variable selection by sipls algorithm

A chemometric-assisted UV absorption spectroscopic method is proposed for the simultaneous determination of acesulfame-K, aspartame and stevioside in raw powder mixtures of commercial sweeteners. The synergy interval partial least squares (siPLS) algorithm was applied to select the optimum spectral range and their combinations. The utilization of spectral region selection aims to construct better partial least squares (PLS) model than that established from the full-spectrum range. The results show that the siPLS algorithm can find out an optimized combination of spectral regions, yielding lower relative standard error of prediction (RSEP) and root mean square error of prediction (RMSEP), as well as simplifying the model. The RMSEP and RSEP obtained after selection of intervals by siPLS were 0.1330 μg·ml–1 and 1.50 % for acesulfame-K, 0.2540 μg·ml–1 and 1.64 % for aspartame, 1.4041 μg·ml–1 and 2.03 % for stevioside respectively. The recovery values range from 98.12 % to 101.88 % for acesulfame-K, 98.63 % to 102.96% for aspartame, and 96.38 % to 104.04 % for stevioside respectively.

A Comparison of Analytical and Data Preprocessing Methods for Spectral Fingerprinting

Spectral fingerprinting, as a method of discriminating between plant cultivars and growing treatments for a common set of broccoli samples, was compared for six analytical instruments. Spectra were acquired for finely powdered solid samples using Fourier transform infrared (FT-IR) and Fourier transform near-infrared (NIR) spectrometry. Spectra were also acquired for unfractionated aqueous methanol extracts of the powders using molecular absorption in the ultraviolet (UV) and visible (VIS) regions and mass spectrometry with negative (MS-) and positive (MS+) ionization. The spectra were analyzed using nested one-way analysis of variance (ANOVA) and principal component analysis (PCA) to statistically evaluate the quality of discrimination. All six methods showed statistically significant differences between the cultivars and treatments. The significance of the statistical tests was improved by the judicious selection of spectral regions (IR and NIR), masses (MS+ and MS-), and derivatives (IR, NIR, UV, and VIS).

Comparison of univariate and multivariate calibration for the determination of micronutrients in pellets of plant materials by laser induced breakdown spectrometry

The application of laser induced breakdown spectrometry (LIBS) aiming the direct analysis of plant materials is a great challenge that still needs efforts for its development and validation. In this way, a series of experimental approaches has been carried out in order to show that LIBS can be used as an alternative method to wet acid digestions based methods for analysis of agricultural and environmental samples. The large amount of information provided by LIBS spectra for these complex samples increases the difficulties for selecting the most appropriated wavelengths for each analyte. Some applications have suggested that improvements in both accuracy and precision can be achieved by the application of multivariate calibration in LIBS data when compared to the univariate regression developed with line emission intensities. In the present work, the performance of univariate and multivariate calibration, based on partial least squares regression (PLSR), was compared for analysis of pellets of plant materials made from an appropriate mixture of cryogenically ground samples with cellulose as the binding agent. The development of a specific PLSR model for each analyte and the selection of spectral regions containing only lines of the analyte of interest were the best conditions for the analysis. In this particular application, these models showed a similar performance, but PLSR seemed to be more robust due to a lower occurrence of outliers in comparison to the univariate method. Data suggests that efforts dealing with sample presentation and fitness of standards for LIBS analysis must be done in order to fulfill the boundary conditions for matrix independent development and validation.

Rapid identification of coagulase-negative staphylococci by Fourier transform infrared spectroscopy

Coagulase-negative staphylococci (CNS), frequently associated with both community-acquired and nosocomial bloodstream infections, must be distinguished from Staphylococcus aureus for clinical purposes. Conventional methods are too laborious and time-consuming and often lack sensitivity to CNS. Fourier transform infrared (FTIR) spectroscopy combined with the use of a universal growth medium (Que-Bact® Universal Medium No. 2) and chemometrics was evaluated for its potential as a rapid and simple clinical tool for making this distinction. FTIR spectra of 11 methicillin-sensitive and 11 methicillin-resistant CNS isolates as well as 25 methicillin-sensitive, 47 methicillin-resistant, 34 borderline oxacillin-resistant and 35 glycopeptide intermediate S. aureus isolates were obtained from dried films of stationary-phase cells grown on the universal medium. Principal component analysis (PCA), self-organizing maps, and the K-nearest neighbor algorithm were employed to cluster the different phenotypes based on similarity of their FTIR spectra. PCA of the first-derivative normalized spectral data from a single narrow region (2888–2868 cm − 1 ) yielded complete differentiation of CNS from both methicillin-sensitive and methicillin-resistant S. aureus. The rate of correct classification was somewhat reduced, from 100% to 90%, after inclusion of borderline oxacillin-resistant and glycopeptide intermediate S. aureus strains in the data set. Differentiation based on the data in broader spectral regions was much less reliable. The results of this study indicate that with proper spectral region selection, FTIR spectroscopy and cluster analysis may provide a simple and accurate means of CNS species identification.

Screening analysis of river seston downstream of an effluent discharge point using near-infrared reflectance spectrometry and wavelet-based spectral region selection

A methodology for screening analysis of river seston downstream of an industry effluent by using near-infrared reflectance spectrometry was developed. A wavelet transform (WT)-based strategy is used to select a spectral region in which the effect of the effluent on the optical properties of the seston is more evident. The methodology was applied to samples from the River Mumbaba in northeast Brazil. Four sites were monitored: two upstream (1 and 2), one at the discharge point of the effluent (3), and another downstream (4). Soft Independent Modelling of Class Analogies (SIMCA) models were built for site 1 and were then applied to the classification of samples from sites 2 and 4. The results reveal that the WT-based spectral region selection is essential to ensure good sensitivity and specificity with respect to the detection of events associated to the effluent discharges at site 3. In fact, the changes in site 4 caused by the effluent are masked by other environmental factors when the full spectrum is employed.

Classification of human gliomas by infrared imaging spectroscopy and chemometric image processing

As a molecular probe of tissue composition, infrared spectroscopic imaging can potentially serve as an adjunct to histopathology in detecting and diagnosing disease. This study demonstrates that human gliomas are distinguishable from control tissue on the basis of IR image used in combination with chemometric imaging processing. Using an iterative two-step algorithm – comprised of a linear discriminant analysis guided genetic optimal spectral region selection – tissue types can be discriminated from one another thus providing insight into the malignancy grade of the tissue. A series of classification models were built using a k-fold cross validation scheme and the classification predictions from the various models were combined to provide an aggregated prediction. The validation of the aggregated model reveals an improvement in the classification success rate to 64%.

The effects of the Arctic haze as determined from airborne radiometric measurements during AGASP II

The interaction of the Aretic winter aerosol (Arctic haze) with solar radiation produces changes in the radiation field that result in the enhancement of scattering and absorption processes which alter the energy balance and solar energy distribution in the Arctic atmosphere-surface system. During the second Arctic Gas and Aerosols Sampling Project (AGASP II) field experiment, we measured radiation parameters using the NOAA WP-3D research aircraft as a platform. State-of-the-art instrumentation was used to measure in situ the absorption of solar radiation by the Arctic atmosphere during severe haze events. Simultaneously with the absorption measurements, we determined optical depths, and total, direct, and scattered radiation fields. All optical measurements were made at spectral bands centered at 412, 500, 675, and 778 nm and with a bandpass of 10 nm. With this selection of spectral regions we concentrated on the measurement of the radiative effects of the aerosol excluding most of the contributions by the gaseous components of the atmosphere. An additional measurement performed during these experiments was the determination of total solar spectrum fluxes. The experimentally determined parameters were used to define an aerosol model that was employed to deduce the absorption by the aerosols over the full solar spectrum and to calculate atmospheric heating rate profiles. The analyses summarized above allowed us to deduce the magnitude of the change in some important parameters. For example, we found changes in instantaneous heating rate of up to about 0.6 K/day. Besides the increased absorption (30 to 40%) and scattering of radiation by the atmosphere, the haze reduces the surface absorption of solar energy by 6 to 10% and the effective planetary albedo over ice surfaces by 3 to 6%. The vertical distribution of the absorbing aerosol is inferred from the flux measurements. Values for the specific absorption of carbon are found to be around 6 m2/g for externally mixed aerosol and about 11.7 m2/g for internally mixed aerosol. A complete study of the radiative effects of the Arctic haze should include infrared measurements and calculations as well as physics of the ice, snow, and water surfaces.