Prediction of the occurrence of leprosy reactions based on Bayesian networks.

Introduction: Leprosy reactions (LR) are severe episodes of intense activation of the host inflammatory response of uncertain etiology, today the leading cause of permanent nerve damage in leprosy patients. Several genetic and non-genetic risk factors for LR have been described; however, there are limited attempts to combine this information to estimate the risk of a leprosy patient developing LR. Here we present an artificial intelligence (AI)-based system that can assess LR risk using clinical, demographic, and genetic data. Methods: The study includes four datasets from different regions of Brazil, totalizing 1,450 leprosy patients followed prospectively for at least 2 years to assess the occurrence of LR. Data mining using WEKA software was performed following a two-step protocol to select the variables included in the AI system, based on Bayesian Networks, and developed using the NETICA software. Results: Analysis of the complete database resulted in a system able to estimate LR risk with 82.7% accuracy, 79.3% sensitivity, and 86.2% specificity. When using only databases for which host genetic information associated with LR was included, the performance increased to 87.7% accuracy, 85.7% sensitivity, and 89.4% specificity. Conclusion: We produced an easy-to-use, online, free-access system that identifies leprosy patients at risk of developing LR. Risk assessment of LR for individual patients may detect candidates for close monitoring, with a potentially positive impact on the prevention of permanent disabilities, the quality of life of the patients, and upon leprosy control programs.

Prediction of the occurrence of leprosy reactions based on Bayesian networks

Introduction: Leprosy reactions (LR) are severe episodes of intense activation of the host inflammatory response of uncertain etiology, today the leading cause of permanent nerve damage in leprosy patients. Several genetic and non-genetic risk factors for LR have been described; however, there are limited attempts to combine this information to estimate the risk of a leprosy patient developing LR. Here we present an artificial intelligence (AI)-based system that can assess LR risk using clinical, demographic, and genetic data. Methods: The study includes four datasets from different regions of Brazil, totalizing 1,450 leprosy patients followed prospectively for at least 2 years to assess the occurrence of LR. Data mining using WEKA software was performed following a two-step protocol to select the variables included in the AI system, based on Bayesian Networks, and developed using the NETICA software. Results: Analysis of the complete database resulted in a system able to estimate LR risk with 82.7% accuracy, 79.3% sensitivity, and 86.2% specificity. When using only databases for which host genetic information associated with LR was included, the performance increased to 87.7% accuracy, 85.7% sensitivity, and 89.4% specificity. Conclusion: We produced an easy-to-use, online, free-access system that identifies leprosy patients at risk of developing LR. Risk assessment of LR for individual patients may detect candidates for close monitoring, with a potentially positive impact on the prevention of permanent disabilities, the quality of life of the patients, and upon leprosy control programs.

[Data mining and characteristics of infant mortality]. / Mineração de dados e características da mortalidade infantil.

This study aims to identify patterns in maternal and fetal characteristics in the prediction of infant mortality by incorporating innovative techniques like data mining, with proven relevance for public health. A database was developed with infant deaths from 2000 to 2004 analyzed by the Committees for the Prevention of Infant Mortality, based on integration of the Information System on Live Births (SINASC), Mortality Information System, and Investigation of Infant Mortality in the State of Paraná. The data mining software was WEKA (open source). The data mining conducts a database search and provides rules to be analyzed to transform the data into useful information. After mining, 4,230 rules were selected: teenage pregnancy plus birth weight < 2,500 g, or post-term birth plus teenage mother with a previous child or intercurrent conditions increase the risk of neonatal death. The results highlight the need for greater attention to teenage mothers, newborns with birth weight < 2,500 g, post-term neonates, and infants of mothers with intercurrent conditions, thus corroborating other studies.

Mineração de dados e características da mortalidade infantil / Data mining and characteristics of infant mortality

O estudo busca identificar padrões de características materno-fetais na predição da mortalidade infantil, por meio da incorporação de técnicas inovadoras, como a Mineração de Dados, que se mostram relevantes em Saúde Pública. Foi elaborada uma base de dados, com óbitos infantis analisados pelos Comitês de Prevenção da Mortalidade Infantil de 2000 a 2004, a partir da integração dos Sistemas de Informações de Nascidos Vivos, da Mortalidade e da Investigação da Mortalidade Infantil no Estado do Paraná. O programa da mineração foi o WEKA, de uso livre. A mineração faz busca em banco de dados e fornece regras que devem ser analisadas para transformação em informação útil. Após a mineração, selecionaram-se 4.230 regras, por exemplo: mãe adolescente e peso ao nascer < 2.500g, ou parto pós-termo e mãe adolescente com outro filho, ou com afecções maternas, aumentam o risco para óbito neonatal. Vê-se a necessidade de estabelecer maior atenção às adolescentes, às crianças com peso ao nascer < 2.500g, pós-termo e filhas de mães com afecções maternas, confirmando resultados de outros estudos.

This study aims to identify patterns in maternal and fetal characteristics in the prediction of infant mortality by incorporating innovative techniques like data mining, with proven relevance for public health. A database was developed with infant deaths from 2000 to 2004 analyzed by the Committees for the Prevention of Infant Mortality, based on integration of the Information System on Live Births (SINASC), Mortality Information System, and Investigation of Infant Mortality in the State of Paraná. The data mining software was WEKA (open source). The data mining conducts a database search and provides rules to be analyzed to transform the data into useful information. After mining, 4,230 rules were selected: teenage pregnancy plus birth weight < 2,500g, or post-term birth plus teenage mother with a previous child or intercurrent conditions increase the risk of neonatal death. The results highlight the need for greater attention to teenage mothers, newborns with birth weight < 2,500g, post-term neonates, and infants of mothers with intercurrent conditions, thus corroborating other studies.

Brain-computer interface using wavelet transformation and naïve bayes classifier.

The main purpose of this work is to establish an exploratory approach using electroencephalographic (EEG) signal, analyzing the patterns in the time-frequency plane. This work also aims to optimize the EEG signal analysis through the improvement of classifiers and, eventually, of the BCI performance. In this paper a novel exploratory approach for data mining of EEG signal based on continuous wavelet transformation (CWT) and wavelet coherence (WC) statistical analysis is introduced and applied. The CWT allows the representation of time-frequency patterns of the signal's information content by WC qualiatative analysis. Results suggest that the proposed methodology is capable of identifying regions in time-frequency spectrum during the specified task of BCI. Furthermore, an example of a region is identified, and the patterns are classified using a Naïve Bayes Classifier (NBC). This innovative characteristic of the process justifies the feasibility of the proposed approach to other data mining applications. It can open new physiologic researches in this field and on non stationary time series analysis.

Sistema de auxílio ao diagnóstico clínico das crises epilépticas / A system to help on the clinical diagnosis of epileptic crises

Apresenta-se um sistema baseado em conhecimentos, para auxiliar no diagnóstico clínico das crises epilépticas, tendo como modêlo a classificação por tipo de crise da International League against Epilepsia. O objetivo do sistema é obter um conjunto de sintomas apresentado pelo paciente, classificar o tipo de crise e indicar o provável diagnóstico...

Estudo da integraçäo de técnicas de multimídia para a construçäo de sistemas tutoriais inteligentes. Projeto TURIM: tutorial em rede, inteligente e multimídia / Study of the multimidia technique integration to contruction of intelligent tutorial systems. TURIM project: tutorial in network, intelligent and multimidia

Os recentes avanços da Informática na área de multimídia permitem visualizar um novo o profícuo campo de pesquisa: a integração destes recursos aos sistemas ICAI. Sabidamente o poder de transmissão do conhecimento por meio de imagens e sons é por vezes superior aos tradicionais métodos de leitura. Considera-se que a integração destes novos meios permitirá um salto qualitativo importante no desenvolvimento dos sistemas ICAI, especialmente em determinadas áreas do conhecimento em que imagem e som são de reconhecida importância. Pode-se citar como uma destas áreas o ensino do diagnóstico médico a partir de imagens (EEG, ECG, ultrasons, etc), considerada como a motivação principal deste trabalho. O projeto pretende fornecer um sistema capaz de integrar o usuário a um meio que permita, por exemplo, auxiliar no aprendizado da cardiologia, fazendo com que este aprendizado seja o mais próximo possível de um tutor real. Para isto utilizará técnicas de inteligência artificial, sistemas distribuídos e multimídia.