Categoria: Artigos Científicos

Resumo: GISSA é um sistema inteligente para a tomada de decisões em saúde focado no cuidado materno infantil. Neste sistema, vários alertas são gerados nos cinco domínios da saúde (clínico-epidemiológico, normativo, administrativo, gestão do conhecimento, conhecimento compartilhado). O sistema se propõe a contribuir para a redução da mortalidade infantil no Brasil. Este artigo apresenta o LAIS, um mecanismo inteligente que usa aprendizado de máquina para gerar alertas de risco de mortalidade infantil no GISSA. Para tanto, este trabalho usa uma metodologia baseada na mineração de dados para alcançar um modelo de aprendizagem capaz de calcular a probabilidade de um recém-nascido morrer. Os testes mostram que o classificador Naive Bayes é o mais adequado para este propósito, apresentando bons resultados, com área da curva ROC de 92,1%. O trabalho reúne bases de dados do Ministério da Saúde, SIM e SINASC, para o treinamento de algoritmos de classificação, identificando relações entre dados de nascimento e de morte de crianças com menos de um an. Durante o processo metodológico foi utilizado o algoritmo spread subsample, que aplica sub-amostragem, melhorando os resultados do modelo.

Resumo: Tomar decisões de boa governança é um desafio constante para a administração da Saúde Pública. Os gestores de saúde precisam fazer análises de dados para identificar diversos problemas de saúde. No Brasil, esses dados são disponibilizados pelo DATASUS. Geralmente, eles são armazenados em bancos de dados distintos e heterogêneos. A abordagem Linked Data permite uma visão homogeneizada dos dados como uma base única. Este artigo propõe um modelo baseado em ontologia e Linked Data para integrar conjuntos de dados e calcular a probabilidade de risco de morte materna e infantil para dar suporte à tomada de decisão no projeto GISSA.

Abstract: In healthcare, uncertainty moments are frequent, especially when they come from diseases with similar signals and symptoms. This work proposes a mobile health application based on predictive classifiers as inference mechanism capable to support health professionals in the identification of diseases transmitted by the Aedes Aegypti mosquito. The proposed system identifies the most probable disease in the case of dengue and chikungunya, given a set of symptoms presented by a patient. This work evaluates the experiments by crossvalidation using real data, and the results show that decision tree perform well for the proposed solution.

Abstract: This paper proposes a better way to represent the architecture of LARIISA, an intelligent system for decision making in healthcare. The proposed representation weaves health and computational domains in a multidimensional architecture, which facilitates the visualization of specialized applications added to the LARIISA framework. New concepts as Big Data, Internet of Things and Linked Data are also introduced to the proposed architecture. Acquiring new data from different sources to the LARIISA database, and making use of it, will permit a more efficient decision-making process for the system.

Abstract: VITESSE is a low cost system to support users in two scenarios: home care and accidents (fainting, trampling, etc.). Initially, the system was based on the digital TV technology in scenarios of home care. Nowadays, the system adds new functions to support urgent and emergency care of individuals in mobility. In both cases, the key idea of VITESSE is to improve the time of consuming process, taking into account the real time and contextual information, in particular in the case of accidents of mobile users. Therefore, VITESSE is a context-aware system that makes use of the concept of Internet of Things (IoT) and ontologies in the process of generating inferences, increasing the efficiency of health care systems.

Abstract: The home care consists in a form of primary care performed by a lay caregiver, a specialist or a multidisciplinary team. This modality is applied in elderly people or patients in treatment of chronic disease who are not at risk of death. The aim of this work is to present a set of context-aware health applications in a prototype of software and hardware that will assist caregivers and/or patients in home care situations. For this, a Set-Top Box (STB) connected to a TV with access to the Internet is used as a way of user interaction, which may enter information about its current state. Furthermore, health sensors can be used to capture data continuously to feed the system. The raw data and information provided by the user are later used, allowing, then, an inference about the patient condition.

Abstract: Health care services can be scarce and expensive in some countries and especially in isolated regions. The lack of information can degrade health care services, for example, by ineffective resource allocation or failure in epidemiological prediction. This paper proposes an architecture for system of decision making and service provisioning in the health care context. It encompasses and integrates data produced by environmental sensors installed in the assisted homes, medical data sets, domainspecific and semantic enriched data sets, and all data generated and collected in applications installed on mobile phones, wearable devices, desktops, web servers, and smart television. LARIISA architecture is presented as a platform to manage, provide and launch services that monitor and analyze data to supply relevant information to decision makers and health care actors that participate in the health care supply chain

Many intrinsically related determinants of health and disease exist, including social and economic status, education, employment, housing, and physical and environmental exposures. These factors interact to cumulatively affect health and disease burden of individuals and populations, and to establish health inequities and disparities across and within countries. Biomedical models of health care decrease adverse consequences of disease, but are not enough to effectively improve individual and population health and advance health equity. Social determinants of health are especially important in Latin American countries, which are characterised by adverse colonial legacies, tremendous social injustice, huge socioeconomic disparities, and wide health inequities.

Abstract: Starting in the late 1980s, many Latin American countries began social sector reforms to alleviate poverty, reduce socioeconomic inequalities, improve health outcomes, and provide financial risk protection. In particular, starting in the 1990s, reforms aimed at strengthening health systems to reduce inequalities in health access and outcomes focused on expansion of universal health coverage, especially for poor citizens. In Latin America, health-system reforms have produced a distinct approach to universal health coverage, underpinned by the principles of equity, solidarity, and collective action to overcome social inequalities. In most of the countries studied, government financing enabled the introduction of supply-side interventions to expand insurance coverage for uninsured citizens—with defined and enlarged benefits packages—and to scale up delivery of health services. Countries such as Brazil and Cuba introduced tax-financed universal health systems. These changes were combined with demand-side interventions aimed at alleviating poverty (targeting many social determinants of health) and improving access of the most disadvantaged populations. Hence, the distinguishing features of health-system strengthening for universal health coverage and lessons from the Latin American experience are relevant for countries advancing universal health coverage.

Abstract: This paper presents improvements of LARIISA, a framework that makes use of context-aware information to support decision-making and governance in the public health area. More specifically, two relevant e-health applications are presented to illustrate the LARIISA system. The first one uses Bayesian networks in dengue scenarios. The second application uses ontology to manage home care scenarios. In both cases, the contributions related to the LARIISA framework include patient health diagnosis provided remotely, support for decision-making health systems, and context information for context-aware health systems.