Novas abordagens para configurações automáticas dos parâmetros de controle em comitês de classificadores

Significant advances have emerged in research related to the topic of Classifier Committees. The models that receive the most attention in the literature are those of the static nature, also known as ensembles. The algorithms that are part of this class, we highlight the methods that using techniques of resampling of the training data: Bagging, Boosting and Multiboosting. The choice of the architecture and base components to be recruited is not a trivial task and has motivated new proposals in an attempt to build such models automatically, and many of them are based on optimization methods. Many of these contributions have not shown satisfactory results when applied to more complex problems with different nature. In contrast, the thesis presented here, proposes three new hybrid approaches for automatic construction for ensembles: Increment of Diversity, Adaptive-fitness Function and Meta-learning for the development of systems for automatic configuration of parameters for models of ensemble. In the first one approach, we propose a solution that combines different diversity techniques in a single conceptual framework, in attempt to achieve higher levels of diversity in ensembles, and with it, the better the performance of such systems. In the second one approach, using a genetic algorithm for automatic design of ensembles. The contribution is to combine the techniques of filter and wrapper adaptively to evolve a better distribution of the feature space to be presented for the components of ensemble. Finally, the last one approach, which proposes new techniques for recommendation of architecture and based components on ensemble, by techniques of traditional meta-learning and multi-label meta-learning. In general, the results are encouraging and corroborate with the thesis that hybrid tools are a powerful solution in building effective ensembles for pattern classification problems.

Decoding Spontaneous Emotional States in the Human Brain.

Pattern classification of human brain activity provides unique insight into the neural underpinnings of diverse mental states. These multivariate tools have recently been used within the field of affective neuroscience to classify distributed patterns of brain activation evoked during emotion induction procedures. Here we assess whether neural models developed to discriminate among distinct emotion categories exhibit predictive validity in the absence of exteroceptive emotional stimulation. In two experiments, we show that spontaneous fluctuations in human resting-state brain activity can be decoded into categories of experience delineating unique emotional states that exhibit spatiotemporal coherence, covary with individual differences in mood and personality traits, and predict on-line, self-reported feelings. These findings validate objective, brain-based models of emotion and show how emotional states dynamically emerge from the activity of separable neural systems.

Towards person-centered neuroimaging markers for resilience and vulnerability in Bipolar Disorder

Improved clinical care for Bipolar Disorder (BD) relies on the identification of diagnostic markers that can reliably detect disease-related signals in clinically heterogeneous populations. At the very least, diagnostic markers should be able to differentiate patients with BD from healthy individuals and from individuals at familial risk for BD who either remain well or develop other psychopathology, most commonly Major Depressive Disorder (MDD). These issues are particularly pertinent to the development of translational applications of neuroimaging as they represent challenges for which clinical observation alone is insufficient. We therefore applied pattern classification to task-based functional magnetic resonance imaging (fMRI) data of the n-back working memory task, to test their predictive value in differentiating patients with BD (n=30) from healthy individuals (n=30) and from patients' relatives who were either diagnosed with MDD (n=30) or were free of any personal lifetime history of psychopathology (n=30). Diagnostic stability in these groups was confirmed with 4-year prospective follow-up. Task-based activation patterns from the fMRI data were analyzed with Gaussian Process Classifiers (GPC), a machine learning approach to detecting multivariate patterns in neuroimaging datasets. Consistent significant classification results were only obtained using data from the 3-back versus 0-back contrast. Using contrast, patients with BD were correctly classified compared to unrelated healthy individuals with an accuracy of 83.5%, sensitivity of 84.6% and specificity of 92.3%. Classification accuracy, sensitivity and specificity when comparing patients with BD to their relatives with MDD, were respectively 73.1%, 53.9% and 94.5%. Classification accuracy, sensitivity and specificity when comparing patients with BD to their healthy relatives were respectively 81.8%, 72.7% and 90.9%. We show that significant individual classification can be achieved using whole brain pattern analysis of task-based working memory fMRI data. The high accuracy and specificity achieved by all three classifiers suggest that multivariate pattern recognition analyses can aid clinicians in the clinical care of BD in situations of true clinical uncertainty regarding the diagnosis and prognosis.

Complicaciones perioperatorias en el manejo de corrección de escoliosis en un hospital de tercer nivel en la ciudad de Bogotá entre el 2010-2014

Introducción: la escoliosis, definida como una deformidad de la columna vertebral en más de 10 grados, se agrupa en 4 orígenes distintos: idiopática, congénita, neuromuscular y sindromática. Cada una de ellas con diferente riesgo de progresión en severidad, lo que determina la necesidad de corrección quirúrgica para cada caso en su tratamiento. Conocer las probabilidades de complicación en la etapa peri operatoria, abre la posibilidad de dar asesoría integral que mida la relación riesgo - beneficio de la medida terapéutica. Métodos: se realiza un estudio retrospectivo de corte transversal. La información se obtiene de los registros de las historias clínicas desde el año 2010 al 2014, de pacientes intervenidos quirúrgicamente para la corrección de escoliosis. Resultados: Se obtuvieron 318 registros de procedimientos en 230 pacientes. El tipo de escoliosis presentado con mayor frecuencia es de origen idiopático 108 (47%); en los 4 tipos de escoliosis se observa mayor número de mujeres 169 (73,4%). La edad donde se concentran la mayor cantidad de cirugías para corrección de escoliosis está entre 10 - 14 años. De 13 complicaciones seleccionadas, aquellas de origen respiratorio son las de mayor probabilidad de ocurrencia (OR 30 - sig 0,000). La característica sociodemográfica “edad” logra predecir el 46% de las complicaciones perioperatorias. Discusión: La corrección de escoliosis va acompañada de comorbilidades, datos sociodemográficos y diagnósticos que en conjunto determinan el grado de complicación peri operatoria. Se necesitan registros clínicos muy completos para poder determinar la asociación entre la etiología de la escoliosis con las complicaciones más comunes. Este trabajo propone y evidencia los datos de los registros clínicos como predictores de complicaciones quirúrgicas de escoliosis. Esto exige un trabajo institucional interno que garantice la calidad en los registros de datos clínicos.

Interactions between landcover pattern and geospatial processing methods:effects on landscape metrics and classification accuracy

Remote sensing data is routinely used in ecology to investigate the relationship between landscape pattern as characterised by land use and land cover maps, and ecological processes. Multiple factors related to the representation of geographic phenomenon have been shown to affect characterisation of landscape pattern resulting in spatial uncertainty. This study investigated the effect of the interaction between landscape spatial pattern and geospatial processing methods statistically; unlike most papers which consider the effect of each factor in isolation only. This is important since data used to calculate landscape metrics typically undergo a series of data abstraction processing tasks and are rarely performed in isolation. The geospatial processing methods tested were the aggregation method and the choice of pixel size used to aggregate data. These were compared to two components of landscape pattern, spatial heterogeneity and the proportion of landcover class area. The interactions and their effect on the final landcover map were described using landscape metrics to measure landscape pattern and classification accuracy (response variables). All landscape metrics and classification accuracy were shown to be affected by both landscape pattern and by processing methods. Large variability in the response of those variables and interactions between the explanatory variables were observed. However, even though interactions occurred, this only affected the magnitude of the difference in landscape metric values. Thus, provided that the same processing methods are used, landscapes should retain their ranking when their landscape metrics are compared. For example, highly fragmented landscapes will always have larger values for the landscape metric "number of patches" than less fragmented landscapes. But the magnitude of difference between the landscapes may change and therefore absolute values of landscape metrics may need to be interpreted with caution. The explanatory variables which had the largest effects were spatial heterogeneity and pixel size. These explanatory variables tended to result in large main effects and large interactions. The high variability in the response variables and the interaction of the explanatory variables indicate it would be difficult to make generalisations about the impact of processing on landscape pattern as only two processing methods were tested and it is likely that untested processing methods will potentially result in even greater spatial uncertainty. © 2013 Elsevier B.V.

Association between overbite and craniofacial growth pattern

The purpose of the present study was to assess the association between overbite and craniofacial growth pattern. The sample comprised eighty-six cephalograms obtained during the orthodontic pretreatment phase and analyzed using the Radiocef program to identify the craniofacial landmarks and perform orthodontic measurements. The variables utilized were overbite, the Jarabak percentage and the Vert index, as well as classifications resulting from the interpretation of these measurements. In all the statistical tests, a significance level of 5% was considered. Measurement reliability was checked by calculating method error. Weighted Kappa analysis showed that agreement between the facial types defined by the Vert index and the direction of growth trend established by the Jarabak percentage was not satisfactory. Owing to this lack of equivalency, a potential association between overbite and craniofacial growth pattern was evaluated using the chi-square test, considering the two methods separately. No relationship of dependence between overbite and craniofacial growth pattern was revealed by the results obtained. Therefore, it can be concluded that the classification of facial growth pattern will not be the same when considering the Jarabak and the Ricketts anayses, and that increased overbite cannot be associated with a braquifacial growth pattern, nor can openbite be associated with a dolichofacial growth pattern.

Mucin pattern reflects the origin of the adenocarcinoma in Barrett's esophagus: a retrospective clinical and laboratorial study

Background: Mucin immunoexpression in adenocarcinoma arising in Barrett's esophagus (BE) may indicate the carcinogenesis pathway. The aim of this study was to evaluate resected specimens of adenocarcinoma in BE for the pattern of mucins and to correlate to the histologic classification. Methods: Specimens were retrospectively collected from thirteen patients who underwent esophageal resection due to adenocarcinoma in BE. Sections were scored for the grade of intestinal metaplasia. The tissues were examined by immunohistochemistry for MUC2 and MUC5AC antibodies. Results: Eleven patients were men. The mean age was 61 years old (varied from 40 to 75 years old). The tumor size had a mean of 4.7 +/- 2.3 cm, and the extension of BE had a mean of 7.7 +/- 1.5 cm. Specialized epithelium with intestinal metaplasia was present in all adjacent mucosas. Immunohistochemistry for MUC2 showed immunoreactivity in goblet cells, while MUC5AC was extensively expressed in the columnar gastric cells, localizing to the surface epithelium and extending to a variable degree into the glandular structures in BE. Tumors were classified according to the mucins in gastric type in 7/13 (MUC5AC positive) and intestinal type in 4/13 (MUC2 positive). Two tumors did not express MUC2 or MUC5AC proteins. The pattern of mucin predominantly expressed in the adjacent epithelium was associated to the mucin expression profile in the tumors, p = 0.047. Conclusion: Barrett's esophagus adenocarcinoma shows either gastric or intestinal type pattern of mucin expression. The two types of tumors developed in Barrett's esophagus may reflect the original cell type involved in the malignant transformation.

DOES THE CHRONOTYPE CLASSIFICATION NEED TO BE UPDATED? PRELIMINARY FINDINGS

Traditionally, chronotype classification is based on the Morningness-Eveningness Questionnaire (MEQ). It is implicit in the classification that intermediate individuals get intermediate scores to most of the MEQ questions. However, a small group of individuals has a different pattern of answers. In some questions, they answer as ""morning-types"" and in some others they answer as ""evening-types,"" resulting in an intermediate total score. ""Evening-type"" and ""Morning-type"" answers were set as A(1) and A(4), respectively. Intermediate answers were set as A(2) and A(3). The following algorithm was applied: Bimodality Index = (Sigma A(1) x Sigma A(4))(2) - (Sigma A(2) x Sigma A(3))(2). Neither-types that had positive bimodality scores were classified as bimodal. If our hypothesis is validated by objective data, an update of chronotype classification will be required. (Author correspondence: brunojm@ymail.com)

A classification methodology for the risk of weed infestation using fuzzy logic

Despite modern weed control practices, weeds continue to be a threat to agricultural production. Considering the variability of weeds, a classification methodology for the risk of infestation in agricultural zones using fuzzy logic is proposed. The inputs for the classification are attributes extracted from estimated maps for weed seed production and weed coverage using kriging and map analysis and from the percentage of surface infested by grass weeds, in order to account for the presence of weed species with a high rate of development and proliferation. The output for the classification predicts the risk of infestation of regions of the field for the next crop. The risk classification methodology described in this paper integrates analysis techniques which may help to reduce costs and improve weed control practices. Results for the risk classification of the infestation in a maize crop field are presented. To illustrate the effectiveness of the proposed system, the risk of infestation over the entire field is checked against the yield loss map estimated by kriging and also with the average yield loss estimated from a hyperbolic model.

Classification of individuals with dyslipidaemia controlled by statins according to plasma biomarkers of oxidative stress using cluster analysis

Oxidative stress is a physiological condition that is associated with atherosclerosis. and it can be influenced by diet. Our objective was to group fifty-seven individuals with dyslipidaemia controlled by statins according to four oxidative biomarkers, and to evaluate the diet pattern and blood biochemistry differences between these groups. Blood samples were collected and the following parameters were evaluated: diet intake; plasma fatty acids; lipoprotein concentration; glucose; oxidised LDL (oxLDL); malondialdehyde (MDA): total antioxidant activity by 2,2-diphenyl-1-picrylhydrazyl (DPPH) and ferric reducing ability power assays. Individuals were separated into five groups by cluster analysis. All groups showed a difference with respect to at least one of the four oxidative stress biomarkers. The separation of individuals in the first axis was based upon their total antioxidant activity. Clusters located on the right side showed higher total antioxidant activity, higher myristic fatty acid and lower arachidonic fatty acid proportions than clusters located on the left side. A negative correlation was observed between DPPH and the peroxidability index. The second axis showed differences in oxidation status as measured by MDA and oxLDL concentrations. Clusters located on the Upper side showed higher oxidative status and lower HDL cholesterol concentration than clusters located on the lower side. There were no differences in diet among the five clusters. Therefore, fatty acid synthesis and HDL cholesterol concentration seem to exert a more significant effect on the oxidative conditions of the individuals with dyslipidaemia controlled by statins than does their food intake.

Centrilobular Fibrosis: An Underrecognized Pattern in Systemic Sclerosis

Background: The impressive association of lung involvement and gastroesophageal reflux in scleroderma raises the possibility of a cause-effect relationship. Objectives: To determine clinical, radiological and histopathological features of systemic sclerosis (SSc) patients according the presence or absence of centrilobular fibrosis (CLF). Methods: Twenty-eight SSc patients with lung involvement were submitted to open lung biopsy and the specimens classified for the presence of CLF (bronchocentric distribution of the lesions and intraluminal matter according to the classification of idiopathic interstitial pneumonia). HRCT, pulmonary function tests and esophageal analysis were also performed. Subsequently, cyclophosphamide was introduced for the nonspecific interstitial pneumonia subgroup and antireflux treatment was intensified for isolated CLF patients. Results: Isolated CLF was found in 21% of the biopsies and also found associated to nonspecific interstitial pneumonia in 84% of these patients. The other 3 cases had usual interstitial pneumonia, pulmonary hypertension and respiratory bronchiolitis-associated interstitial lung disease. The histopathological analysis revealed that all 6 patients with isolated CLF had the bronchocentric distribution and intraluminal basophilic content, with foreign bodies detected in one third of them. The central distribution of lung involvement on HRCT was found in 67% of these patients with a consistent patchy distribution (100%). Ground glass (67%) and consolidation (33%) were the predominant patterns found. The constant clinical finding in all isolated CLF cases was dyspnea, esophageal abnormalities and a moderate lung impairment (FVC: 63.83 +/- 16.31%; DLCO: 61.66 +/- 18.84%). Lung function parameters in isolated CLF patients remained stable after 1 year of exclusively intensive antireflux treatment (FVC, p = 0.23; DLCO, p = 0.59). Conclusions: The novel description of CLF pattern in SSc lung disease with peculiar histological, tomographic and clinical features will certainly contribute to a more appropriate therapeutic approach. Copyright (C) 2008 S. Karger AG, Basel

Classification System of Raman Spectra using Cluster Analysis to Diagnose Coronary Artery Lesions

The traditional methods employed to detect atherosclerotic lesions allow for the identification of lesions; however, they do not provide specific characterization of the lesion`s biochemistry. Currently, Raman spectroscopy techniques are widely used as a characterization method for unknown substances, which makes this technique very important for detecting atherosclerotic lesions. The spectral interpretation is based on the analysis of frequency peaks present in the signal; however, spectra obtained from the same substance can show peaks slightly different and these differences make difficult the creation of an automatic method for spectral signal analysis. This paper presents a signal analysis method based on a clustering technique that allows for the classification of spectra as well as the inference of a diagnosis about the arterial wall condition. The objective is to develop a computational tool that is able to create clusters of spectra according to the arterial wall state and, after data collection, to allow for the classification of a specific spectrum into its correct cluster.

Human visceral leishmaniasis expresses Th1 pattern in situ liver lesions

The architectural and infiltrate pattern of liver human visceral leishmaniasis (HVL) have been systematically classified as typical, fibrogenic or nodular. Despite this histopathological classification, the immune response based on cytokines and cellular phenotypes have never been performed. The aim of this study was to determine the immunophenotypic pattern and cytokine profile of the nodular involvement of the Liver in HVL. We evaluated nine cases of the nodular form of HVL. In situ immune response was studied through cytokine analysis and immunohistochemical study for phenotype markers: IL-1, IL-4, IL-1 0, TNF-alpha, IFN-gamma, CD4+ T cells, CD8+ T cells, CD20, CD68, CD57 and macrophage activation was determined by evaluation of iNOS activity. HVL seems to be related to a better immune response. Amastigotes were rarely found on liver sections. Leishmania antigen expression was also rare and located in the inflammatory nodules. The lower expression of IL-4 and IL-10, moderate expression of TNF-alpha and IFN-gamma demonstrate a panorama of Th1 phenotype. The increased expression of NK cells could help in sustaining this model of response. This pattern of immune response is probably responsible for improvement in the parasite`s clearance from liver tissue and it is a prognostic marker of human visceral leishmaniasis. (C) 2008 The British Infection Society. Published by Elsevier Ltd. All rights reserved.

Diffuse liver disease classification from ultrasound surface characterization, clinical and laboratorial data

In this work liver contour is semi-automatically segmented and quantified in order to help the identification and diagnosis of diffuse liver disease. The features extracted from the liver contour are jointly used with clinical and laboratorial data in the staging process. The classification results of a support vector machine, a Bayesian and a k-nearest neighbor classifier are compared. A population of 88 patients at five different stages of diffuse liver disease and a leave-one-out cross-validation strategy are used in the classification process. The best results are obtained using the k-nearest neighbor classifier, with an overall accuracy of 80.68%. The good performance of the proposed method shows a reliable indicator that can improve the information in the staging of diffuse liver disease.

Fatty liver characterization and classification by ultrasound

Steatosis, also known as fatty liver, corresponds to an abnormal retention of lipids within the hepatic cells and reflects an impairment of the normal processes of synthesis and elimination of fat. Several causes may lead to this condition, namely obesity, diabetes, or alcoholism. In this paper an automatic classification algorithm is proposed for the diagnosis of the liver steatosis from ultrasound images. The features are selected in order to catch the same characteristics used by the physicians in the diagnosis of the disease based on visual inspection of the ultrasound images. The algorithm, designed in a Bayesian framework, computes two images: i) a despeckled one, containing the anatomic and echogenic information of the liver, and ii) an image containing only the speckle used to compute the textural features. These images are computed from the estimated RF signal generated by the ultrasound probe where the dynamic range compression performed by the equipment is taken into account. A Bayes classifier, trained with data manually classified by expert clinicians and used as ground truth, reaches an overall accuracy of 95% and a 100% of sensitivity. The main novelties of the method are the estimations of the RF and speckle images which make it possible to accurately compute textural features of the liver parenchyma relevant for the diagnosis.