EEG signal classification for epilepsy diagnosis via optimum path forest - A systematic assessment

Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)

Classificação de tábuas de madeira usando processamento de imagens digitais e aprendizado de máquina

Pós-graduação em Agronomia (Energia na Agricultura) - FCA

Estimating the concentration of gold nanoparticles incorporated on natural rubber membranes using multi-level starlet optimal segmentation

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

Automatic classification of prostate stromal tissue in histological images using Haralick descriptors and Local Binary Patterns

In this paper we presente a classification system that uses a combination of texture features from stromal regions: Haralick features and Local Binary Patterns (LBP) in wavelet domain. The system has five steps for classification of the tissues. First, the stromal regions were detected and extracted using segmentation techniques based on thresholding and RGB colour space. Second, the Wavelet decomposition was applied in the extracted regions to obtain the Wavelet coefficients. Third, the Haralick and LBP features were extracted from the coefficients. Fourth, relevant features were selected using the ANOVA statistical method. The classication (fifth step) was performed with Radial Basis Function (RBF) networks. The system was tested in 105 prostate images, which were divided into three groups of 35 images: normal, hyperplastic and cancerous. The system performance was evaluated using the area under the ROC curve and resulted in 0.98 for normal versus cancer, 0.95 for hyperplasia versus cancer and 0.96 for normal versus hyperplasia. Our results suggest that texture features can be used as discriminators for stromal tissues prostate images. Furthermore, the system was effective to classify prostate images, specially the hyperplastic class which is the most difficult type in diagnosis and prognosis.

Desenvolvimento de um sistema inteligente para o diagnóstico e prognóstico de falhas em sistemas de distribuição de energia elétrica

Pós-graduação em Engenharia Elétrica - FEIS

Deteccão de complexos QRS em eletrocardiogramas baseada na decomposição em valores singulares em multirresolução

Pós-graduação em Engenharia Elétrica - FEIS

Segmentação de texturas em imagens digitais utilizando wavelets redundantes

Pós-graduação em Matematica Aplicada e Computacional - FCT

Mid- to late-Holocene paleoceanographic changes on the southeastern Brazilian shelf based on grain size records

; High-resolution grain size analyses of three AMS (14)C-dated cores from the Southeastern Brazilian shelf provide a detailed record of mid- to late-Holocene environmental changes in the Southwestern Atlantic Margin. The cores exhibit millennial variability that we associate with the previously described southward shift of the Inter Tropical Convergence Zone (ITCZ) average latitudinal position over the South American continent during the Holocene climatic maximum. This generated changes in the wind-driven current system of the SW Atlantic margin and modified the grain size characteristics of the sediments deposited there. Centennial variations in the grain size are associated with a previously described late-Holocene enhancement of the El Nino-Southern Oscillation (ENSO) amplitude, which led to stronger NNE trade winds off eastern Brazil, favouring SW transport of sediments from the Paraiba do Sul River. This is recorded in a core from off Cabo Frio as a coarsening trend from 3000 cal. BP onwards. The ENSO enhancement also caused changes in precipitation and wind pattern in southern Brazil, allowing high discharge events and northward extensions of the low-saline water plume from Rio de la Plata. We propose that this resulted in a net increase in northward alongshore transport of fine sediments, seen as a prominent fine-shift at 2000 cal. BP in a core from similar to 24 degrees S on the Brazilian shelf. Wavelet-and spectral analysis of the sortable silt records show a significant similar to 1000-yr periodicity, which we attribute to solar forcing. If correct, this is one of the first indications of solar forcing of this timescale on the Southwestern Atlantic margin.

Multidecadal climate variability in Brazil's Nordeste during the last 3000 years based on speleothem isotope records

We present the first high resolution, approximately similar to 4 years sample spacing, precipitation record from northeastern Brazil (hereafter referred to as 'Nordeste') covering the last similar to 3000 yrs from Th-230-dated stalagmites oxygen isotope records. Our record shows abrupt fluctuations in rainfall tied to variations in the intensity of the South American summer monsoon (SASM), including the periods corresponding to the Little Ice Age (LIA), the Medieval Climate Anomaly (MCA) and an event around 2800 yr B.P. Unlike other monsoon records in southern tropical South America, dry conditions prevailed during the LIA in the Nordeste. Our record suggests that the region is currently undergoing drought conditions that are unprecedented over the past 3 millennia, rivaled only by the LIA period. Using spectral, wavelet and cross-wavelet analyses we show that changes in SASM activity in the region are mainly associated with variations of the Atlantic Multidecadal Oscillation (AMO) and to a lesser degree caused by fluctuations in tropical Pacific SST. Our record also shows a distinct periodicity around 210 years, which has been linked to solar variability. Citation: Novello, V. F., et al. (2012), Multidecadal climate variability in Brazil's Nordeste during the last 3000 years based on speleothem isotope records, Geophys. Res. Lett., 39, L23706, doi: 10.1029/2012GL053936.

A novel method for power quality multiple disturbance decomposition based on Independent Component Analysis

In this paper, a novel method for power quality signal decomposition is proposed based on Independent Component Analysis (ICA). This method aims to decompose the power system signal (voltage or current) into components that can provide more specific information about the different disturbances which are occurring simultaneously during a multiple disturbance situation. The ICA is originally a multichannel technique. However, the method proposes its use to blindly separate out disturbances existing in a single measured signal (single channel). Therefore, a preprocessing step for the ICA is proposed using a filter bank. The proposed method was applied to synthetic data, simulated data, as well as actual power system signals, showing a very good performance. A comparison with the decomposition provided by the Discrete Wavelet Transform shows that the proposed method presented better decoupling for the analyzed data. (C) 2012 Elsevier Ltd. All rights reserved.

Comparing non-stationary and irregularly spaced time series

In this paper, we present approximate distributions for the ratio of the cumulative wavelet periodograms considering stationary and non-stationary time series generated from independent Gaussian processes. We also adapt an existing procedure to use this statistic and its approximate distribution in order to test if two regularly or irregularly spaced time series are realizations of the same generating process. Simulation studies show good size and power properties for the test statistic. An application with financial microdata illustrates the test usefulness. We conclude advocating the use of these approximate distributions instead of the ones obtained through randomizations, mainly in the case of irregular time series. (C) 2012 Elsevier B.V. All rights reserved.

Solar-Terrestrial Signal Record in Tree Ring Width Time Series from Brazil

This work investigates the behavior of the sunspot number and Southern Oscillation Index (SOI) signal recorded in the tree ring time series for three different locations in Brazil: Humaita in Amaznia State, Porto Ferreira in So Paulo State, and Passo Fundo in Rio Grande do Sul State, using wavelet and cross-wavelet analysis techniques. The wavelet spectra of tree ring time series showed signs of 11 and 22 years, possibly related to the solar activity, and periods of 2-8 years, possibly related to El Nio events. The cross-wavelet spectra for all tree ring time series from Brazil present a significant response to the 11-year solar cycle in the time interval between 1921 to after 1981. These tree ring time series still have a response to the second harmonic of the solar cycle (5.5 years), but in different time intervals. The cross-wavelet maps also showed that the relationship between the SOI x tree ring time series is more intense, for oscillation in the range of 4-8 years.

Variability of rainfall and temperature (1912-2008) parameters measured from Santa Maria (29 degrees 41 ' S, 53 degrees 48 ' W) and their connections with ENSO and solar activity

In this work, we analyze the long term variability of rainfall and temperature (1912-2008) of Santa Maria (29 degrees S, 53 degrees W) and its possible connection with natural influences such as solar activity and ENSO. Temperature and rainfall present similar frequencies as revealed by spectral analyses. This analysis shows a large number of short periods between 2-8 years and periods of 11.8-12.3, 19.1-21.0, and 64.3-82.5 years. The cross correlation for rainfall and temperature versus Southern Oscillation Index (SOI) have higher cross-power around 2-8 yr. Rainfall and temperature versus sunspot number (Rz) showed higher cross-power around the 11-yr solar cycle period. A high and continuous cross correlation was observed for Rz-22 yr versus rainfall and temperature. Furthermore, the power between 22-yr solar cycle and meteorological parameters was higher than that obtained with the 11-yr solar cycle, suggesting that the effect of Hale cycle on climate may be stronger than the Schwabe cycle effect. These results indicate that the variability of rainfall and temperature is closely related to the variation of the Southern Oscillation Index and solar activity, and that the El Nino Southern Oscillation and solar activity probably play an important role in the climate system over Southern Brazil. (C) 2011 Elsevier Ltd. All rights reserved.

A comparative study on multiscale fractal dimension descriptors

Fractal theory presents a large number of applications to image and signal analysis. Although the fractal dimension can be used as an image object descriptor, a multiscale approach, such as multiscale fractal dimension (MFD), increases the amount of information extracted from an object. MFD provides a curve which describes object complexity along the scale. However, this curve presents much redundant information, which could be discarded without loss in performance. Thus, it is necessary the use of a descriptor technique to analyze this curve and also to reduce the dimensionality of these data by selecting its meaningful descriptors. This paper shows a comparative study among different techniques for MFD descriptors generation. It compares the use of well-known and state-of-the-art descriptors, such as Fourier, Wavelet, Polynomial Approximation (PA), Functional Data Analysis (FDA), Principal Component Analysis (PCA), Symbolic Aggregate Approximation (SAX), kernel PCA, Independent Component Analysis (ICA), geometrical and statistical features. The descriptors are evaluated in a classification experiment using Linear Discriminant Analysis over the descriptors computed from MFD curves from two data sets: generic shapes and rotated fish contours. Results indicate that PCA, FDA, PA and Wavelet Approximation provide the best MFD descriptors for recognition and classification tasks. (C) 2012 Elsevier B.V. All rights reserved.

Automatic lumen segmentation in IVOCT images using binary morphological reconstruction

Abstract Background Atherosclerosis causes millions of deaths, annually yielding billions in expenses round the world. Intravascular Optical Coherence Tomography (IVOCT) is a medical imaging modality, which displays high resolution images of coronary cross-section. Nonetheless, quantitative information can only be obtained with segmentation; consequently, more adequate diagnostics, therapies and interventions can be provided. Since it is a relatively new modality, many different segmentation methods, available in the literature for other modalities, could be successfully applied to IVOCT images, improving accuracies and uses. Method An automatic lumen segmentation approach, based on Wavelet Transform and Mathematical Morphology, is presented. The methodology is divided into three main parts. First, the preprocessing stage attenuates and enhances undesirable and important information, respectively. Second, in the feature extraction block, wavelet is associated with an adapted version of Otsu threshold; hence, tissue information is discriminated and binarized. Finally, binary morphological reconstruction improves the binary information and constructs the binary lumen object. Results The evaluation was carried out by segmenting 290 challenging images from human and pig coronaries, and rabbit iliac arteries; the outcomes were compared with the gold standards made by experts. The resultant accuracy was obtained: True Positive (%) = 99.29 ± 2.96, False Positive (%) = 3.69 ± 2.88, False Negative (%) = 0.71 ± 2.96, Max False Positive Distance (mm) = 0.1 ± 0.07, Max False Negative Distance (mm) = 0.06 ± 0.1. Conclusions In conclusion, by segmenting a number of IVOCT images with various features, the proposed technique showed to be robust and more accurate than published studies; in addition, the method is completely automatic, providing a new tool for IVOCT segmentation.