Edilson Delgado-Trejos

Dimensionality Reduction Oriented Toward the Feature Visualization for Ischemia Detection

An effective data representation methodology on high-dimension feature spaces is presented, which allows a better interpretation of subjacent physiological phenomena (namely, cardiac behavior related to cardiovascular diseases), and is based on search criteria over a feature set resulting in an increase in the detection capability of ischemic pathologies, but also connecting these features with the physiologic representation of the ECG. The proposed dimension reduction scheme consists of three levels: projection, interpretation, and visualization. First, a hybrid algorithm is described that projects the multidimensional data to a lower dimension space, gathering the features that contribute similarly in the meaning of the covariance reconstruction in order to find information of clinical relevance over the initial training space. Next, an algorithm of variable selection is provided that further reduces the dimension, taking into account only the variables that offer greater class separability, and finally, the selected feature set is projected to a 2-D space in order to verify the performance of the suggested dimension reduction algorithm in terms of the discrimination capability for ischemia detection. The ECG recordings used in this study are from the European ST-T database and from the Universidad Nacional de Colombia database. In both cases, over 99% feature reduction was obtained, and classification precision was over 99% using a five-nearest-neighbor classifier (5-NN).

Nonlinear Dynamics Techniques for the Detection of the Brain Areas Using MER Signals

A methodology for identifying brain areas from the brain MER signals (microelectrode recordings) is presented, which is based on a nonlinear feature set. We propose nonlinear dynamics measures such as correlation dimension, Hurst exponent and the largest Lyapunov exponent to characterize the dynamic structure. The MER records belong to the Polytechnical University of Valencia, 24 records for each zone (black substance, thalamus, subthalamus nucleus and uncertain area). The detection of each area using characteristics derived from complexity analysis was obtained through a classifier (support vector machine). The joint information between areas is remarkable and the best accuracy result was 93.75%. The nonlinear dynamics techniques help to discriminate the four brain areas considered, since they take into account the intrinsic dynamics of the signals and the structures analysis based on the multivariate statistical procedures is an important step in the data preprocessing.

Método para el Diagnóstico de Rodamientos Utilizando la Complejidad de Lempel-Ziv

La presencia de una falla en un rodamiento hace que el sistema mecanico evolucione de una dinamica debilmente no lineal a una dinamica fuertemente no lineal, por lo tanto los metodos lineales comunes en el dominio del tiempo y la frecuencia no son adecuados para el diagnostico de rodamientos. En el presente articulo se propone una metodologia novedosa no lineal para la deteccion de fallas en rodamientos, que usa la medida de complejidad sugerida por Lempel y Ziv para caracterizar las senales de vibracion. La ventaja principal de este metodo sobre las demas tecnicas de analisis no lineal es que no requiere la reconstruccion de un atractor, por lo que es adecuado para realizar analisis en tiempo real. Los resultados obtenidos muestran que la complejidad de Lempel-Ziv es una herramienta efectiva para el diagnostico de rodamientos.

On detecting determinism and nonlinearity in microelectrode recording signals: Approach based on non-stationary surrogate data methods

Two new surrogate methods, the Small Shuffle Surrogate (SSS) and the Truncated Fourier Transform Surrogate (TFTS), have been proposed to study whether there are some kind of dynamics in irregular fluctuations and if so whether these dynamics are linear or not, even if this fluctuations are modulated by long term trends. This situation is theoretically incompatible with the assumption underlying previously proposed surrogate methods. We apply the SSS and TFTS methods to microelectrode recording (MER) signals from different brain areas, in order to acquire a deeper understanding of them. Through our methodology we conclude that the irregular fluctuations in MER signals possess some determinism.

Clasificador No Lineal Basado en Redes Neuronales con Funciones de Base Radial para Implementación en Sistemas de Punto Fijo

Implementation of intelligent machines requires of efficient classification systems under limited computational resources. Thisstudy introduces a method for estimating the parameters of Radial Basis Function Neural Network (RBF-NN) that can be implemented on a fixed point processor. First, the number of hidden nodes is chosen based on statistics of the mapped data points. A k-means search is then carried out to determine the location of each node. The hidden units mapping corresponds to the Euclidean distance of their centers to each data point, the weights of the output sum are obtained by solving a linear least squares problem. With this procedure, a low computational cost classifier can be readily implemented on a low capacity platform for real time applications.

Localization of superficially buried objects by seismic-acoustic techniques

This paper describes the procedure for obtaining ground prospections using seismic-acoustic signals, oriented to the detection of buried objects located at the ground surface. The presented methodology uses a Helmholtz resonator as the source of perturbation and a set of omnidirectional microphones used as sensors for detecting reflections of the superficial Rayleigh waves. The results show that the methodology is able to provide ground images where the buried object can be easily detected, both in plain as well as in uneven surfaces.

Dynamic signature for a closed-set identification based on nonlinear analysis

This paper presents a study of biometric identification using a methodology based on complexity measures. The identification system designed, implemented and evaluated uses nonlinear dynamic techniques such as Lempel-Ziv Complexity, the Largest Lyapunov Exponent, Hurst Exponent, Correlation Dimension, Shannon Entropy and Kolmogorov Entropy to characterize the process and capture the intrinsic dynamics of the users signature. In the validation process 3 databases were used SVC, MCYT and our own (ITMMS-01) obtaining closed-set identification performances of 98.12%, 97.38% and 99.50% accordingly. Satisfactory results were achieved with a conventional linear classifier spending a minimum computational cost.

Implementacion en Sistemas Embebidos de la Transformada Discreta de Hartley

Orthogonal transformations have been very useful in the characterization and signal processing. In particular Hartley Transform allows for time-frequency representations and vice versa. This paper presents an algorithm for calculating the Discrete Hartley Transform in embedded systems with the objective of minimizing the computational load and storage capacity required. It exploits the similarity with the Discrete Fourier Transform to use a fast algorithm reduces the number of trigonometric functions calculated using the rotation factors (Twiddle factors). In general, the implementation can increase the size of the processing window and increase computational speed compared to direct calculation.

Análisis de la Tolerancia al Ruido de Características Basadas en Dinámica no Lineal Sobre Señales Fonocardiográficas

A very desirable attribute in automatic pathology detection systems is noise robustness, such that the presence of noise should not significantly affect the ability to detect pathologies. This paper explores the discriminatory potential that nonlinear dynamic features, in particular, the Hurst exponent, the maximum Lyapunov exponent and correlation dimension, can provide in the detection of heart murmurs using phonocardiographic signals contaminated with different levels of noise. At the same time comparing with features obtained from time frequency representations. The results show the strength of features based on nonlinear dynamics for functional state classification tasks, even on signals with high noise levels.

Hypernasal Speech Detection by Acoustic Analysis of Unvoiced Plosive Consonants

People with a defective velopharyngeal mechanism speak with abnormal nasal resonance (hypernasal speech). Voice analysis methods for hypernasality detection commonly use vowels and nasalized vowels. However to obtain a more general assessment of this abnormality it is necessary to analyze stops and fricatives. This study describes a method with high generalization capability for hypernasality detection analyzing unvoiced Spanish stop consonants. The importance of phoneme-by-phoneme analysis is shown, in contrast with whole word parametrization which includes irrelevant segments from the classification point of view. Parameters that correlate the imprints of Velopharyngeal Incompetence (VPI) over voiceless stop consonants were used in the feature estimation stage. Classification was carried out using a Support Vector Machine (SVM), including the Rademacher complexity model with the aim of increasing the generalization capability. Performances of 95.2% and 92.7% were obtained in the processing and verification stages for a repeated cross-validation classifier evaluation.