Complementary use of partial least-squares and artificial neural networks for the non-linear spectrophotometric analysis of pharmaceutical samples

The complementary use of partial least-squares (PLS) multivariate calibration and artificial neural networks (ANNs) for the simultaneous spectrophotometric determination of three active components in a pharmaceutical formulation has been explored. The presence of non-linearities caused by chemical interactions was confirmed by a recently discussed methodology based on Mallows augmented partial residual plots. Ternary mixtures of chlorpheniramine, naphazoline and dexamethasone in a matrix of excipients have been resolved by using PLS for the two major analytes (chlorpheniramine and naphazoline) and ANNs for the minor one (dexamethasone). Notwithstanding the large number of constituents, their high degree of spectral overlap and the occurrence of non-linearities, rapid and simultaneous analysis has been achieved, with reasonably good accuracy and precision. No extraction procedures using non-aqueous solvents are required.     

Determination of physical properties of bitumens by use of near-infrared spectroscopy with neural networks. Joint modelling of linear and non-linear parameters

The fact that bitumens behave as non-Newtonian fluids results in non-linear relationships between their near-infrared (NIR) spectra and the physico-chemical properties that define their consistency (viz. penetration and viscosity). Determining such properties using linear calibration techniques [e.g. partial least-squares regression (PLSR)] entails the previous transformation of the original variables by use of non-linear functions and employing the transformed variables to construct the models. Other properties of bitumens such as density and composition exhibit linear relationships with their NIR spectra. Artificial neural networks (ANNs) enable modelling of systems with a non-linear property-spectrum relationship; also, they allow one to determine several properties of a sample with a single model, so they are effective alternatives to linear calibration methods. In this work, the ability of ANNs simultaneously to determine both linear and non-linear parameters for bitumens without the need previously to transform the original variables was assessed. Based on the results, ANNs allow the simultaneous determination of several linear and non-linear physical properties typical of bitumens.     

人工神经元网络法提高毛细管电泳药物控制分析准确度的研究

研究了人工神经元网络法在毛细管电泳定量测定memantine中提高测定准确度 的可行性。在毛细管电泳法定量测定memantine的过程中，其浓度与峰高或峰面积 以及与二者和内标的比值均没有良好的线性关系。人工神经元网络具有很强的非线 性校正能力，其最大优点是无须对分离体系及组分的迁移行为预先予以了解。人工 神经元网络的输为memantine的峰高和峰面积，输出为memantine的浓度。通过实验 确定的网络结构为2：1：1型。由于人工神经元网络的通用性，该法也可用于毛细 管电泳在其他药物控制分析中改善定量分析的准确度。     

Handling complex effects in slurry-sampling-electrothermal atomic absorption spectrometry by multivariate calibration

Analysis of solid samples by slurry-sampling-electrothermal atomic absorption spectrometry (SS-ETAAS) can imply spectral and chemical interferences caused by the large amount of concomitants introduced into the graphite furnace. Sometimes they cannot be solved using stabilized temperature platform furnace (STPF) conditions or typical approaches (previous sample ashing, use of chemical modifiers, etc.), which are time consuming and quite expensive. A new approach to handle interferences using multivariate calibrations (partial least squares, PLS, and artificial neural networks, ANN) is presented and exemplified with a real problem consisting on determining Sb in several solid matrices (soils, sediments and coal fly ash) as slurries by ETAAS.Experimental designs were implemented at different levels of Sb to develop the calibration matrix and assess which concomitants (seven ions were considered) modified the atomic signal mostly. They were Na⁺ and Ca²⁺ and they induced simultaneous displacement, depletion (enhancement) and broadening of the atomic peak. Here it is shown that these complex effects can be handled in a reliable, fast and cost-effective way to predict the concentration of Sb in slurry samples of several solid matrices. The method was validated predicting the concentrations of five certified reference materials (CRMs) and studying its robustness to current ETAAS problems. It is also shown that linear PLS can handle eventual non-linearities and that its results are comparable to more complex (non-linear) models, as those from ANNs.     

Artificial neural networks for computer-aided modelling and optimisation in micellar electrokinetic chromatography

The separation process in capillary micellar electrochromatography (MEKC) can be modelled using artificial neural networks (ANNs) and optimisation of MEKC methods can be facilitated by combining ANNs with experimental design. ANNs have shown attractive possibilities for non-linear modelling of response surfaces in MEKC and it was demonstrated that by combining ANN modelling with experimental design, the number of experiments necessary to search and find optimal separation conditions can be reduced significantly. A new general approach for computer-aided optimisation in MEKC has been proposed which, because of its general validity, can also be applied in other separation techniques.     

Enhanced application of square wave voltammetry with glassy carbon electrode coupled to multivariate calibration tools for the determination of B(6) and B(12) vitamins in pharmaceutical preparations

Vitamins B₆ (VB₆) and B₁₂ (VB₁₂) were simultaneously determined in pharmaceutical preparations by using square wave voltametry (SWV) together with artificial neural networks (ANNs). Supporting electrolyte solution, pH and voltametric technique were optimised. The calibration set was built with several artificial samples containing both active ingredients and excipients. Deviations from linearity were observed for both analytes. It is probably due to interactions among the electro active components and competition by the electrode surface, fact that supports the use of ANNs. Recoveries when analysing a nine sample validation set, of 100.2 and 96.4 were calculated for VB₆ and VB₁₂, respectively. Commercial samples were analysed with reasonably good results considering the complexity of the mixture studied.     

A review of multivariate calibration methods applied to biomedical analysis

The determination of the contents of therapeutic drugs, metabolites and other important biomedical analytes in biological samples is usually performed by using high-performance liquid chromatography (HPLC). Modern multivariate calibration methods constitute an attractive alternative, even when they are applied to intrinsically unselective spectroscopic or electrochemical signals. First-order (i.e., vectorized) data are conveniently analyzed with classical chemometric tools such as partial least-squares (PLS). Certain analytical problems require more sophisticated models, such as artificial neural networks (ANNs), which are especially able to cope with non-linearities in the data structure. Finally, models based on the acquisition and processing of second- or higher-order data (i.e., matrices or higher dimensional data arrays) present the phenomenon known as “second-order advantage”, which permits quantitation of calibrated analytes in the presence of interferents. The latter models show immense potentialities in the field of biomedical analysis. Pertinent literature examples are reviewed.     

Flow injection analysis of fluoride: optimization of experimental conditions and non-linear calibration using artificial neural networks

This paper deals with the application of artificial neural networks (ANNs) to two common problems in spectroscopy: optimization of experimental conditions and non-linear calibration of the result, with particular reference to the determination of fluoride by flow injection analysis (FIA). The FIA system was based on the formation of a blue ternary complex between zirconium(IV), p-methyldibromoarsenazo and F- with the maximum absorption wavelength at 635 nm. First, optimization in terms of sensitivity and sampling rate was carried out by using jointly a central composite design and ANNs, and a neural network with a 3-7-1 structure was confirmed to be able to provide the maximum performance. Second, the relationship between the concentration of fluoride and its absorbance was modeled by ANNs. In this process, cross-validation and leave-k-out were used. The results showed that good prediction was attained in the 1-4-1 neural net. The trained networks proved to be very powerful in both applications. The proposed method was successfully applied to the determination of free fluoride in tea and toothpaste with recoveries between 96 and 101%.     

Synergism between constituents of multicomponent catalysts designed for ethanol steam reforming using partial least squares regression and artificial neural networks

Effects of different catalyst components on the catalytic performance in steam reforming of ethanol have been investigated by means of Artificial Neural Networks (ANNs) and Partial Least Square regression (PLSR). The data base consisted of ca. 400 items (catalysts with varied composition), which were obtained from a former catalyst optimization procedure. Marten's uncertainty (jackknife) test showed that simultaneous addition of Ni and Co has crucial effect on the hydrogen production. The catalyst containing both Ni and Co provided remarkable hydrogen production at 450°C. The addition of Ceas modifier to the bimetallic NiCo catalyst has high importance at lower temperatures: the hydrogen concentration is doubled at 350°C. Addition of Pt had only little effect on the product distribution. The outliers in the data set have been investigated by means of Hotelling T2 control chart. Compositions containing high amount of Cu or Ce have been identified as outliers, which points to the nonlinear effect of Cu and Ce on the catalytic performance. ANNs were used for analysis of the non-linear effects: an optimum was found with increasing amount of Cu and Ce in the catalyst composition. Hydrogen production can be improved by Ce only in the absence of Zn. Additionally, negative cross-effect was evidenced between Ni and Cu. The above relationships have been visualized in Holographic Maps, too. Although predictive ability of PLSR is somewhat worse than that of ANN, PLSR provided indirect evidence that ANNs were trained adequately.     

Monitoring substrate and products in a bioprocess with FTIR spectroscopy coupled to artificial neural networks enhanced with a genetic-algorithm-based method for wavelength selection

An experiment was developed as a simple alternative to existing analytical methods for the simultaneous quantitation of glucose (substrate) and glucuronic acid (main product) in the bioprocesses Kombucha by using FTIR spectroscopy coupled to multivariate calibration (partial least-squares, PLS-1 and artificial neural networks, ANNs). Wavelength selection through a novel ranked regions genetic algorithm (RRGA) was used to enhance the predictive ability of the chemometric models. Acceptable results were obtained by using the ANNs models considering the complexity of the sample and the speediness and simplicity of the method. The accuracy on the glucuronic acid determination was calculated by analysing spiked real fermentation samples (recoveries ca. 115%).     

Comparative determination of phosphate and silicate using molybdenum blue by radial basis function and feed-forward neural networks assisted by principal component analysis.

In the present study, chemometric analysis of visible spectral data of phospho-and silico-molybdenum blue complexes was used to develop artificial neural networks (ANNs) for the simultaneous determination of the phosphate and silicate. Combinations of principal component analysis (PCA) with feed-forward neural networks (FFNNs) and radial basis function networks (RBFNs) were built and investigated. The structures of the models were simplified by using the corresponding important principal components as input instead of the original spectra. Number of inputs and hidden nodes, learning rate, transfer functions and number of epochs and SPREAD values were optimized. Performances of methods were tested with root mean square errors prediction (RMSEP, %), using synthetic solutions. The obtained satisfactory results indicate the applicability of this ANN approach based on PCA input selection for determination in highly spectral overlapping. The results obtained by FFNNs and by RBF networks were compared. The applicability of methods was investigated for synthetic samples, for detergent formulations, and for a river water sample.     

EIS multianalyte sensing with an automated SIA system—An electronic tongue employing the impedimetric signal

In this work, the simultaneous quantification of three alkaline ions (potassium, sodium and ammonium) from a single impedance spectrum is presented. For this purpose, a generic ionophore - dibenzo-18-crown-6 - was used as a recognition element, entrapped into a polymeric matrix of polypyrrole generated by electropolymerization. Electrochemical impedance spectroscopy (EIS) and artificial neural networks (ANNs) were employed to obtain and process the data, respectively. In fact, EIS detected the ions exchanged between the medium and the sensing layer whereas ANNs, after an appropriated training process, could turn the impedance spectrum into concentrations values. A sequential injection analysis (SIA) system was employed for operation and to automatically generate the information required for the training of the ANN. Best results were obtained by using a backpropagation neural network made up by two hidden layers: the first one contained three neurons with the radbas transfer function and the second one ten neurons with the tansig transfer function. Three commercial fertilizers were tested employing the proposed methodology on account of the high complexity of their matrix. The experimental results were compared with reference methods.     

On the characterization of novel biologically active steroids: Selection of lipophilicity models of newly synthesized steroidal derivatives by classical and non-parametric ranking approaches

In this paper, the guidelines for the interpretation of the results of quantitative structure-retention relationship (QSRR) modeling, comparison and assessment of the established models, as well as the selection of the best and most consistent QSRR model were presented. Various linear and non-linear chemometric regression techniques were used to build QSRR models for chromatographic lipophilicity prediction of a series of triazole, tetrazole, toluenesulfonylhydrazide, nitrile, dinitrile and dione steroid derivatives. Linear regression (LR) and multiple linear regression (MLR) were used as linear techniques, while artificial neural networks (ANNs) were applied as non-linear modeling techniques. Generated models were statistically evaluated applying different approaches for model comparison and ranking. Two non-parametric methods (generalized pair correlation method – GPCM and sum of ranking differences – SRD) were used for model ranking and assessment of the best model for chromatographic lipophilicity prediction using experimentally obtained logk values and row average as a reference ranking. Both, GPCM and SRD, provided highly similar model choice regardless on a different background. These results are in agreement with the classical approach.     

反向传播人工神经网络用于薄层色谱中氨基酸的结构与保留关系的研究

在测定、收集和计算出一组氨基酸的拓扑指数和各种理化参数之后,再通过相关分析选择其中最有代表性的几个参数作为反向传播人工神经网络的输入参数,用于正相薄层色谱中氨基酸保留规律的研究。结果表明,氨基酸的色谱保留值与其结构之间呈现较强的非线性关系,采用人工神经网络方法比用多元线性回归方法能够更精确地描述这种关系。     

Resolution of Differential Pulse Voltammetric Peaks Using Genetic Algorithm Based Variable Selection‐Partial Least Squares and Principal Component‐Artificial Neural Networks

Differential Pulse Voltammetry has been used for the simultaneous determination of cysteine, tyrosine and trptophan on the unmodified glassy carbon electrode. In the analysis of these analytes in the same samples, the main difficulty is the high degree of overlapping of voltammograms. The relationships between the currents and the concentrations are complex and highly nonlinear. The predictive ability of principal component regression (PCR), partial least squares regression (PLS), genetic algorithm‐partial least squares regression (GA‐PLS) and principal component‐artificial neural networks (PC‐ANNs) were examined for simultaneous determination of three amino acids. For a regression model, everything that could not help in constructing the model may be considered as noise without further specification. PC‐ANN and GA‐PLS use significant data and show superiority over other applied multivariate methods. The proposed method was also applied satisfactorily to determination of analytes in some synthetic samples.     

Simultaneous determination of three fluoroquinolones by linear sweep stripping voltammetry with the aid of chemometrics

A linear sweep stripping voltammetric (LSSV) method has been researched and developed for simultaneous quantitative determination of mixtures of three antibiotic drugs, ofloxacin, norfloxacin and ciprofloxacin. It relies on reductive reaction of the antibiotics at a mercury electrode in a Britton-Robinson buffer (pH 3.78). The voltammograms of these three compounds overlap strongly, and show non-linear character. Thus, it is difficult to analyse the compounds individually in their mixtures. In this work, chemometrics methods such as classical least squares (CLS), principal component regression (PCR), partial least squares (PLS) and radial basis function-artificial neural networks (RBF-ANN) were applied for the simultaneous determination of these compounds. The prediction performance of the calibration models constructed on the basis of these methods was compared. It was shown that satisfactory quantitative results were obtained with the use of the RBF-ANN calibration model relative prediction error (RPE_T) of 8.1% and an average recovery of 101%. This method is able to accommodate non-linear data quite well. The proposed analytical method based on LSSV was applied for the analysis of ofloxacin, norfloxacin and ciprofloxacin antibiotics in bird feedstuffs and their spiked samples, as well as in eye drops with satisfactory results.     

Simultaneous determination of methylene blue and new methylene blue by slab optical waveguide spectroscopy and artificial neural networks

A slab optical waveguide (SOWG) has been used for study of adsorption of both methylene blue (MB) and new methylene blue (NMB) in liquid-solid interface. Adsorption characteristics of MB and NMB on both bare SOWG and silanized SOWG by octadecyltrichlorosilane (ODS) were compared. The simultaneous determinations of both MB and NMB were explored by flow injection SOWG spectrophotometric analysis and artificial neural networks (ANNs) for the first time. Concentrations of MB and NMB were estimated simultaneously with the ANNs. Results obtained with SOWG were compared with those got by conventional UV-visible spectrophotometry.     

Spectrophotometric determination of sorbic and benzoic acids in fruit juices by a net analyte signal-based method with selection of the wavelength range to avoid non-modelled interferences

Sorbic (SOR) and benzoic (BEN) acids were determined in fruit juice samples by using a net analyte signal-based methodology named HLA/GO (an hybrid linear analysis presented by Goicoechea and Olivieri) applied to spectroscopic signals. The calibration set was built with several fruit juices in order to take into account the natural variability and concentrations of both analytes covering the range usually present in commercial samples. Relative errors of prediction (REP %) of 3.6 and 5.2% were calculated for SOR and BEN respectively. Several figures of merit were calculated-sensitivity, selectivity, analytical sensitivity, and limit of detection. The method is quantitative, with reasonably good recoveries and excellent precision (less than 1%). Wavelength selection was applied, based on the concept of net analyte signal regression, and it allowed us to improve the method performance in samples containing non-modelled interferences, e.g. fruit juices different to those used to build the calibration model.     

Simultaneous determination of P,Cl, K and Ca in several control serums by alpha-particle activation analysis using the internal standard method

In order to demonstrate the usefulness of alpha-particle activation analysis, simultaneous determination of P, Cl, K and Ca in commercially available control serums has been studied fundamentally. After thick target yield curves of radionuclides produced from the element to be determined were measured as a function of alpha energy together with those of the interferences, an optimum working standard for the present experiments was provided by applying the internal standard method to a human serum under the most suitable bombardment conditions. Then, the concentrations of the above four elements in several control serums were determined efficiently and reasonably by ordinary alpha-particle activation analysis.     

Application of PARAFAC to a two-component system exhibiting fluorescence resonance energy transfer: from theoretical prediction to experimental validation

One of the conventionally accepted requirements for parallel factor analysis (PARAFAC) to handle the fluorescence excitation emission matrices (EEMs) is the independence of each component's absorption and emission spectra in simple mixtures of fluorophores. EEMs of samples in which F?rster resonance energy transfer (FRET) occurs between fluorophores seem to fail to meet this requirement. A rigorous theoretical treatment of the steady-state kinetics in the present work indicates that the fluorescence in the presence of FRET, excited by relatively weak excitation light intensity, can be reasonably separated into additive contributions from three parts: donors, acceptors and FRET. This prediction is for the first time verified experimentally in sodium dodecyl sulfate micellar solutions containing biphenyl as the energy donor and 2,5-diphenyloxazole as the energy acceptor. The experimental EEMs were well fitted to three components as predicted. A well accepted diagnostic test called core consistency (CC), specifically designed for modeling simple mixtures of fluorophores with PARAFAC, was found to be negative for the 3-component model in the present study. The simultaneous occurrence of good model fit and significantly negative CC when modeling fluorophore mixtures by conventional PARAFAC would be indicative of the presence of physical/chemical processes (e.g., FRET) that deviate from the conventional working requirements for PARAFAC. The extent of FRET has been independently measured or calculated by three methods: 1) decrease in steady state fluorescence of donor; 2) lifetime measurements with population analysis; and 3) Poisson statistics based on PARAFAC-determined distribution constants. The results of the three methods are consistent. The normalized scores of the three components found by PARAFAC also agree to within a few percent with relative concentrations in aqueous and micelle phases determined from distribution constants for the solutions prepared with nine different combinations of total donor and acceptor concentrations. Our theoretical treatment also for the first time spells out in detail the relationship between the PARAFAC scores and concentrations of components, in terms of photophysical constants of the components and spectral shape factors.