Juan Manuel Gutiérrez

Determination of total polyphenol index in wines employing a voltammetric electronic tongue

This work reports the application of a voltammetric electronic tongue system (ET) made from an array of modified graphite-epoxy composites plus a gold microelectrode in the qualitative and quantitative analysis of polyphenols found in wine. Wine samples were analyzed using cyclic voltammetry without any sample pretreatment. The obtained responses were preprocessed employing discrete wavelet transform (DWT) in order to compress and extract significant features from the voltammetric signals, and the obtained approximation coefficients fed a multivariate calibration method (artificial neural network-ANN-or partial least squares-PLS-) which accomplished the quantification of total polyphenol content. External test subset samples results were compared with the ones obtained with the Folin-Ciocalteu (FC) method and UV absorbance polyphenol index (I(280)) as reference values, with highly significant correlation coefficients of 0.979 and 0.963 in the range from 50 to 2400 mg L(-1) gallic acid equivalents, respectively. In a separate experiment, qualitative discrimination of different polyphenols found in wine was also assessed by principal component analysis (PCA).

A review of the use of the potentiometric electronic tongue in the monitoring of environmental systems

This paper introduces electronic tongue systems for remote environmental monitoring applications. This new approach in the chemical sensor field consists of the use of an array of non-specific sensors coupled with a multivariate calibration tool which may form a node of a sensor network. In our work, the proposed arrays were made up of potentiometric sensors based on polymeric membranes, and the subsequent cross-response processing was based on a multilayer artificial neural network model. Two cases are described: the environmental monitoring of ammonium pollutant plus alkaline ions at different measuring sites in the states of Mexico and Hidalgo (Mexico), and the monitoring of heavy metals (Cu^2^+, Pb^2^+, Zn^2^+ and Cd^2^+) in open air waste streams and rivers heading down the Gulf of Mexico.

Data Compression for a Voltammetric Electronic Tongue Modelled with Artificial Neural Networks

Abstract In the study of voltammetric electronic tongues, a key point is the preprocessing of the departure information, the voltammograms which form the response of the sensor array, prior to classification or modeling with advanced chemometric tools. This work demonstrates the use of the discrete wavelet transform (DWT) for compacting these voltammograms prior to modeling. After compression, a system based on artificial neural networks (ANNs) was used for the quantification of the electroactive substances present, using the obtained wavelet decomposition coefficients as their inputs. The Daubechies wavelet of fourth order permitted an effective compression up to 16 coefficients, reducing the original dimension by ca. 10 times. The case studied is a mixture of three oxidizable amino acids:tryptophan, cysteine, and tyrosine. With the reduced information, one ANN per specie was trained using the Bayesian regularization algorithm. The proposed procedure was compared with the more conventional treatments of downsampling the voltammogram, or its feature extraction employing principal component analysis prior to ANNs.

Wavelet neural networks to resolve the overlapping signal in the voltammetric determination of phenolic compounds

Three phenolic compounds, i.e. phenol, catechol and 4-acetamidophenol, were simultaneously determined by voltammetric detection of its oxidation reaction at the surface of an epoxy-graphite transducer. Because of strong signal overlapping, Wavelet Neural Networks (WNN) were used in data treatment, in a combination of chemometrics and electrochemical sensors, already known as the electronic tongue concept. To facilitate calibration, a set of samples (concentration of each phenol ranging from 0.25 to 2.5mM) was prepared automatically by employing a Sequential Injection System. Phenolic compounds could be resolved with good prediction ability, showing correlation coefficients greater than 0.929 when the obtained values were compared with those expected for a set of samples not employed for training.

Remote environmental monitoring employing a potentiometric electronic tongue

This work investigates the use of electronic tongues for environmental monitoring. Electronic tongues were based on arrays of potentiometric sensors plus a complex data processing by artificial neural networks and data transmission by radiofrequency. A first application, intended for a system simulating real conditions in surface water, performed a simultaneous monitoring of ammonium, potassium, sodium, chloride, and nitrate ions. The proposed system allowed us to assess the effect of natural biodegradation stages for these species. A second application was used to monitor concentrations of ammonium, potassium, and sodium in the ‘Ignacio Ramírez’ dam (Mexico). The electronic tongue used here allowed us to determine the content of the three cations in real water samples, although a high matrix effect was encountered for sodium determination. The implemented radio transmission worked robustly during all the experiments, thus demonstrating the viability of the proposed systems for automated remote applications.

SIA system employing the transient response from a potentiometric sensor array: Correction of a saline matrix effect

A Sequential Injection Analysis (SIA) system and an 8-potentiometric all-solid-state sensor array were coupled in a simple and automated electronic tongue device. The potentiometric sensors used were planar microfabricated structures with standard PVC membranes deposited onto a gold contact. The SIA system permitted the automated operation and generation of the calibration data, needed to build an Artificial Neural Network model, thanks to the precise dosing and mixing of volumes of stock solutions. The resolution of a four-ion mixture, i.e. ammonium, sodium, nitrate and chloride was the study case used for characterization of the system. Two different variants for signal acquisition, steady-state and transient recording, were arranged and compared. The dynamic treatment is shown to offer improved performance thanks to the benefits of the kinetic resolution. For this, it first extracts meaningful data from a FFT transform of each sensors transient, which is then fed to an ANN model for estimation of each concentration in the four-ion mixture. While in a standard laboratory situation there was no difference between the two approaches, the dynamic treatment allowed the correction of a matrix effect in the case study, where an uncontrolled saline effect could be counterbalanced.

Voltammetric Electronic Tongue and Support Vector Machines for Identification of Selected Features in Mexican Coffee

This paper describes a new method based on a voltammetric electronic tongue (ET) for the recognition of distinctive features in coffee samples. An ET was directly applied to different samples from the main Mexican coffee regions without any pretreatment before the analysis. The resulting electrochemical information was modeled with two different mathematical tools, namely Linear Discriminant Analysis (LDA) and Support Vector Machines (SVM). Growing conditions (i.e., organic or non-organic practices and altitude of crops) were considered for a first classification. LDA results showed an average discrimination rate of 88% ± 6.53% while SVM successfully accomplished an overall accuracy of 96.4% ± 3.50% for the same task. A second classification based on geographical origin of samples was carried out. Results showed an overall accuracy of 87.5% ± 7.79% for LDA and a superior performance of 97.5% ± 3.22% for SVM. Given the complexity of coffee samples, the high accuracy percentages achieved by ET coupled with SVM in both classification problems suggested a potential applicability of ET in the assessment of selected coffee features with a simpler and faster methodology along with a null sample pretreatment. In addition, the proposed method can be applied to authentication assessment while improving cost, time and accuracy of the general procedure.

EMG pattern recognition using Support Vector Machines classifier for myoelectric control purposes

The present work reports the use of Support Vector Machines (SVMs) as classifier of myoelectric signals. This tool was recently used to analyze data and recognize patterns, but just a few studies report its use in myoelectric registers. The aim of this research is analyze and compare some classification schemes employing Artificial Neural Networks and Linear Discriminant Analysis in order to establish the benefits of SVMs models in pattern recognition tasks. The departure information consists in an Electromyographic (EMG) data base of 12 subjects considering 4 degrees of freedom. Before building interpretation models, a pre-processing stage was done to obtain either autoregressive or frequency domain features.

Spatially Adaptive Regularization Image Restoration Using a Modified Hopfield Network

In this paper we present a technique for localized image regularization using a modified Hopfield neural network (MHNN). The algorithm forms a segmented map of the image and classifies it into several clusters, or regions, and assigns each region a regularization parameter according to its local statistics and the prior knowledge about the image obtained by a Bayesian minimum risk (BMR) restoration method. The image segmentation is performed over the BMR restored image. First, the user selects arbitrarily at least one region, and makes a subjective decision to choose the best estimate from among a set of restored images with different regularization parameter applied to the user-selected region. Then, using this decision the algorithm sets up a perception-based selection of the different regularization parameters for restoring in an adaptive fashion the whole image employing the MHNN computations.

Chocolate Classification by an Electronic Nose with Pressure Controlled Generated Stimulation

In this work, we will analyze the response of a Metal Oxide Gas Sensor (MOGS) array to a flow controlled stimulus generated in a pressure controlled canister produced by a homemade olfactometer to build an E-nose. The built E-nose is capable of chocolate identification between the 26 analyzed chocolate bar samples and four features recognition (chocolate type, extra ingredient, sweetener and expiration date status). The data analysis tools used were Principal Components Analysis (PCA) and Artificial Neural Networks (ANNs). The chocolate identification E-nose average classification rate was of 81.3% with 0.99 accuracy (Acc), 0.86 precision (Prc), 0.84 sensitivity (Sen) and 0.99 specificity (Spe) for test. The chocolate feature recognition E-nose gives a classification rate of 85.36% with 0.96 Acc, 0.86 Prc, 0.85 Sen and 0.96 Spe. In addition, a preliminary sample aging analysis was made. The results prove the pressure controlled generated stimulus is reliable for this type of studies.