Search

Este artigo tem por objetivo reportar os resultados de um mapeamento da utilização dos indicadores de gestão e sua relevância para a atividade de gestão de coordenadores de curso de graduação de Instituições Comunitárias de Ensino Superior. Como metodologia foi realizada uma pesquisa exploratória e descritiva, os dados foram coletados por meio de um questionário enviado aos coordenadores de curso de 10 instituições do COMUNG. Entre os principais resultados encontrados destaca-se a maior relevância dos indicadores das dimensões “Estratégicas e Administrativas”, “Estratégicas” e “Pedagógicas” em relação às dimensões “Administrativas”, “Administrativos e Pedagógicos” e “Financeiras”, corroborando com outros estudos, onde os coordenadores de curso de graduação apresentam maior preocupação com questões pedagógicas e de legislação em detrimento das atividades administrativos, financeiras de seus cursos.

Com a o aumento da disponibilidade de dados, sobretudo no contexto educacional, a Mineração de Dados Educacionais (MDE) tem se tornado cada vez mais importante para a tomada de decisão neste contexto. Um dos principais objetivos da MDE é a predição do desempenho, pois quando se sabe de forma antecipada sobre a performance dos alunos, é possível intervir evitando reprovações, e até mesmo a evasão. Nesse sentido, esse estudo tem como objetivo realizar a previsão do desempenho de alunos, em um conjunto de dados públicos, utilizando algoritmos de Regressão, ademais indicar quais os principais atributos preditores para o desempenho dos alunos. Para isso foi implementado um processo de MDE baseado em 4 etapas descritas por Aggarwal (2015). Como resultado foi identificado que para os dois conjuntos de dados analisados a Árvores de Decisão foi o mais preciso, com uma acurácia de 90% para a disciplina de Matemática, e o Random Forest teve o melhor desempenho para os dados referentes a disciplina de Português, 80% de precisão. Além disso, constatou-se que atributos relacionados as atividades escolares são mais preditores para o desempenho dos alunos, todavia alguns atributos decorrentes de características demográficas e socioeconômicas, também influenciam no desempenho.

Analítica Acadêmica é uma área relativamente nova que traz conceitos de análise e business intelligence para o ensino superior. Este artigo tem por objetivo reportar os resultados de um estudo piloto da aplicação de um modelo de Analítica Acadêmica em um Instituição Comunitária de Ensino Superior. O método de pesquisa foi a experimental e utilizada a análise de conteúdo como metodologia. Os resultados são parciais, mas apresentam-se promissores, destacamos a otimização do tempo, qualidade nas informações apresentadas e a boa avaliação dos coordenadores em relação ao apoio a tomada de decisão, auxiliando o processo de gestão.

Este trabalho apresenta a avaliação de modelos preditivos para a identificação de alunos com risco de reprovação em disciplinas específicas. Para tanto, é utilizado como atributo preditor o percurso curricular realizado previamente pelo aluno antes de cursar uma determinada disciplina. O impacto da utilização de técnicas de balanceamento de carga nas métricas de avaliação dos modelos preditivos é investigado. Os resultados destacaram os melhores desempenhos para os algoritmos Random Forest, J48 e IBK, apresentando uma Acurácia de 71% a 81% e Recall de 75% a 93%, refletindo uma significativa melhoria ao se utilizar a técnica de sobreamostragem SMOTE para o balanceamento de carga.

O crescente interesse em pesquisas sobre sistemas de recomendações educacionais tem motivado o surgimento de novas técnicas e modelos nos últimos anos. Entretanto, as informações existentes sobre a diversidade de mecanismos utilizados para a produção de recomendações no contexto educacional são limitadas. Diante disso, este artigo apresenta uma revisão sistemática da literatura que sintetiza o conhecimento disponível sobre a forma que os recomendadores educacionais produzem as recomendações. Para tal foram selecionados 20 artigos publicados entre 2015 e 2019 de 517 publicações científicas identificadas. Os resultados fornecem conclusões sobre como os recomendadores educacionais funcionam apresentando um panorama das técnicas, entradas e saídas desses sistemas nas pesquisas mais recentes., The growing interest in educational recommender systems has motivated the emergence of new techniques and models in recent years. Despite this, there is limited information on a variety of mechanisms used by such systems to produce recommendations in the educational context. Therefore, this paper presents a systematic literature review that summarizes the available knowledge on the operation of educational recommender systems. Through the execution of the systematic review, 20 research papers published between 2015 and 2019 were selected from an initial set of 517 studies. The results provide findings regarding how educational recommenders work by presenting a panorama of the techniques, inputs and outputs of these systems in the most recent research., Facultad de Informática

The main objective of this study was to develop a freely available mobile software application and education platform in health and safety for aquaculture workers and managers. The application, called AquaSafe, was created in Portuguese and English for the Android system using the Java 8 programming language and the Android Studio development environment. AquaSafe content focusses on the identification of hazards and health risks, hazardous sources and preventive measures for occupational health and safety in the aquaculture sector. It has three predominantly interactive user interfaces to reinforce the main messages, which include games in the form of compliance checklists, quizzes and question-answer tools. The software is designed to act as a platform for assessment and management of occupational hazards, guide decision making on simple techniques or measures to prevent injuries, diseases and fatalities during aquaculture activities. It is envisaged that the AquaSafe mobile software will contribute to the prevention of occupational injuries and diseases in aquaculture. O principal objetivo deste estudo foi desenvolver um aplicativo de educação profissional disponível gratuitamente em saúde e segurança para stakeholders de aquicultura. O aplicativo, chamado AquaSafe, foi criado em português e inglês, em sistema Android usando a linguagem de programação Java 8 e o ambiente de desenvolvimento Android Studio. O conteúdo do AquaSafe concentra-se na identificação de perigos e riscos para a saúde e medidas preventivas para saúde e segurança ocupacional no setor de aquicultura. Possui três interfaces de usuário predominantemente interativas para reforçar as principais mensagens, que incluem listas de verificação de conformidade, questionários e ferramentas de resposta a perguntas. O software foi projetado para atuar como uma plataforma para avaliação e gerenciamento de riscos ocupacionais, orientar a tomada de decisões técnicas ou medidas simples para evitar lesões, doenças e fatalidades durante as atividades laborais. AquaSafe pode ser uma ferramenta para a prevenção de lesões e doenças ocupacionais na aquicultura.

RESUMO Objetivo Este estudo visa conhecer o cenário atual da atuação fonoaudiológica em serviços de TAN no Brasil, identificando seus obstáculos e perspectivas, bem como verificar a adequação dos Programas de TAN nacionais aos indicadores de qualidade pré-estabelecidos. Método Estudo observacional analítico, realizado com fonoaudiólogos no exercício da TAN no Brasil, entre agosto de 2018 e agosto de 2019, por meio de um questionário online estruturado. Foram realizadas análises descritivas e correlacionais dos dados no programa SPSS versão 22.0. Resultados A prática efetiva da TAN entre os profissionais não foi condizente, na sua totalidade, com os protocolos oficiais. 48,5% dos fonoaudiólogos afirmaram interrupção da TAN em algum momento nos locais de trabalho, principalmente pela necessidade de reparo do equipamento (64,7%). Quanto aos registros do fluxo e dados de atendimento, que contemplam os indicadores de qualidade, houve maior controle do número total de neonatos que realizaram a TAN (87,9%) e menor controle quanto aos resultados falsos positivos (21,2%). 81,8% dos fonoaudiólogos afirmaram ter disponibilidade para utilizar um sistema para registro e controle de dados da TAN. Conclusão Apesar de as práticas dos profissionais estarem em geral condizentes com protocolos oficiais, a não linearidade do processo e a falta de controle dos dados constituem obstáculos importantes para a qualidade dos serviços de TAN. A maioria dos programas de TAN nacionais apresentados não cumprem os indicadores de qualidade pré-estabelecidos. Considera-se que a informatização dos registros poderá beneficiar os profissionais e potencializar a implementação da TAN como prevista na legislação e políticas públicas nacionais. ABSTRACT Purpose This study aims to know the current scenario of speech audiology therapy activities at NHS in Brazil, identifying its obstacles and perspectives, as well as verifying the adequacy of national NHS Programs to the pre-established quality indicators. Methods Analytical observational study, carried out with speech therapists in the exercise of NHS in Brazil, between August 2018 and August 2019, through a structured online questionnaire. Descriptive and correlational analyzes of the data were performed using the SPSS version 22.0 program. Results The effective practice of NHS was not entirely consistent with official protocols. 48.5% of speech therapists stated that NHS interruption at some point in the workplace, especially due to the need to repair the equipment (64.7%). As for the flow records and care-related data, which include quality indicators there was greater control over the total number of neonates who underwent NHS (87.9%) and less control over false-positive results (21.2%). 81.8% of speech-language-hearing therapists said they were available to use a system for recording and controlling NHS data. Conclusion Although professionals’ practices are generally consistent with official protocols, the non-linearity of the process and the lack of data control are important obstacles to the quality of NHS services. Most of the national NHS programs presented do not meet the pre-established quality indicators. It is considered that the computerization of records can benefit professionals and enhance the implementation of NHS provided for in Brazilian laws and public policies.

Recommender systems have become one of the main tools for personalized content filtering in the educational domain. Those who support teaching and learning activities, particularly, have gained increasing attention in the past years. This growing interest has motivated the emergence of new approaches and models in the field, in spite of it, there is a gap in literature about the current trends on how recommendations have been produced, how recommenders have been evaluated as well as what are the research limitations and opportunities for advancement in the field. In this regard, this paper reports the main findings of a systematic literature review covering these four dimensions. The study is based on the analysis of a set of primary studies (The online version contains supplementary material available at 10.1007/s10639-022-11341-9.

INTRODUCTION: Hepatocellular carcinoma is the prevalent primary liver cancer, a silent disease that killed 782,000 worldwide in 2018. Multimodal deep learning is the application of deep learning techniques, fusing more than one data modality as the model’s input. PURPOSE: A computer-aided diagnosis system for hepatocellular carcinoma developed with multimodal deep learning approaches could use multiple data modalities as recommended by clinical guidelines, and enhance the robustness and the value of the second-opinion given to physicians. This article describes the process of creation and evaluation of an algorithm for computer-aided diagnosis of hepatocellular carcinoma developed with multimodal deep learning techniques fusing preprocessed computed-tomography images with structured data from patient Electronic Health Records. RESULTS: The classification performance achieved by the proposed algorithm in the test dataset was: accuracy = 86.9%, precision = 89.6%, recall = 86.9% and F-Score = 86.7%. These classification performance metrics are closer to the state-of-the-art in this area and were achieved with data modalities which are cheaper than traditional Magnetic Resonance Imaging approaches, enabling the use of the proposed algorithm by low and mid-sized healthcare institutions. CONCLUSION: The classification performance achieved with the multimodal deep learning algorithm is higher than human specialists diagnostic performance using only CT for diagnosis. Even though the results are promising, the multimodal deep learning architecture used for hepatocellular carcinoma prediction needs more training and test processes using different datasets before the use of the proposed algorithm by physicians in real healthcare routines. The additional training aims to confirm the classification performance achieved and enhance the model’s robustness.

A análise de sentimento é uma área de Mineração de Dados que envolve processamento de linguagem natural, extração de informações, inteligência artificial e aprendizado de máquina. Assim, por meio dessa técnica e, também, das emoções dos alunos em ambientes virtuais de aprendizagem, é possível descobrir padrões e verificar eventuais deficiências durante o processo de aprendizagem. O objetivo deste artigo é apresentar os resultados de um mapeamento sistemático da literatura realizado sobre técnicas, métodos, algoritmos, bibliotecas e ferramentas de mineração de dados educacionais utilizados para análise de sentimentos e emoções dos estudantes em ambientes virtuais de aprendizagem. Além disso, foram identificadas as finalidades para a análise de sentimentos e os tipos de emoções consideradas. Portanto, foram selecionados 20 estudos primários para verificação em profundidade. Os resultados apresentam a predominância de algoritmos de aprendizado de máquina para análise de sentimentos, abordando avaliação de cursos e professores, a eficácia do ambiente de aprendizagem, a satisfação e as dificuldades dos alunos. Além disso, a maioria dos estudos explora a polaridade de sentimento: positivo, negativo e neutro. Palavras-chave: Mineração de dados educacionais. Sentimentos. Emoções. Ambiente virtual de aprendizagem.

Considerando os desafios e os benefícios inerentes à Internet das Coisas (Internet of Things - IoT) para idosos na área da saúde, o estudo teve como objetivo realizar um mapeamento sistemático da literatura, na qual a busca foi realizada em dezembro de 2021, nas bases de dados da Biblioteca Virtual em Saúde (BVS), Web of Science, Pubmed, Embase e Medline. Foram utilizados os descritores: tecnologia, aplicação, saúde, Internet das Coisas e idosos. Foram identificados 873 estudos, e depois de aplicados os critérios de seleção, 39 publicações alcançaram o nível de qualidade para inclusão e síntese. Os artigos revisados apresentam dados significativos sobre a evolução dos tratamentos tecnológicos remotos para os idosos. Este estudo é recomendado para auxiliar em pesquisas que visem aumentar o conhecimento, o uso e a qualidade das tecnologias IoT na área da saúde.

A partir de um mapeamento sistemático buscou-se verificar as contribuições de Learning Analytics e Mineração de Dados Educacionais no contexto educacional brasileiro. Optouse por fontes de busca em três revistas na área de Informática na Educação e Anais de dois eventos de relevância nacional onde foram verificados 136 artigos entre janeiro de 2008 e março de 2020. Após a aplicação dos critérios de exclusão e de qualidade foram selecionados 71 artigos. Os resultados apresentam a maior ocorrência de projetos com finalidade de analisar desempenho acadêmico e prevenção de evasão escolar, embora nos últimos anos os assuntos estejam apresentando diversificação temática. Grande parte são voltados ao Ensino Superior e na modalidade de Educação a Distância. Há variedade de tecnologias e recursos utilizados no desenvolvimento de soluções em LA ressaltando o uso de Linguagem R, MySQL e a ferramenta Weka. From a systematic mapping, Learning Analytics (LA) and Educational Data Mining (EDM) contributions in the Brazilian educational context were verified. The sources chosen were three journals in the area of Computers in Education and annals of events of national relevance where 136 articles between 2008 and 2020 were verified. After applying exclusion and quality criteria, 71 articles were selected. Results show the highest occurrence of projects with the purpose of analyzing academic performance and prevention of school dropout, although in recent years the subjects are presenting thematic diversification. Most of them are focused on Higher and Distance Education. There are a variety of technologies and resources used in the development of solutions in Learning Analytics highlighting the use of R Language, MySQL and Weka tool.

The aim of the present study was to analyze the relationship between the student’s profile and academic achievement, using data of the Family Health Specialization course at the UNA-SUS/UFCSPA (FHS). A total of 4,438 students enrolled in the course until 2018, and all that completed the activities (n = 4219) were considered in the analysis. The characterization of academic performance, the student’ status at the end of the course were considered, and the students were classified as: approved, reproved, evaded or unsubscribed; also the student's final grade was considered to characterize their academic achievement. A descriptive analysis and logistic regression models were constructed to verify an association between the student profiles and academic achievements. Most students were female (63.7%), Brazilian (63.9%), unmarried (66.9%), part of the Mais Médicos Program (54.5%), have less than one year formed (32.0%), had never performed distance education activities (72.5%), had internet access at home (94.9%) and at work (64.5%), and used the computer at least six days a week (53.4%). Among the students, 2,643 (52.6%) approved and 1,576 (47,4%) reproved or did not finish the cours. Gender, course entry form, undergraduate course, age at the beginning of the course and region of professional activity were associated with the achievement in the course and grade in complex cases essay. Keywords: Academic achievement. Distance learning. e-larning. Primary health care. O objetivo do presente trabalho foi estudar a relação entre as características do aluno do curso de Especialização em Saúde da Família da UNA-SUS/UFCSPA (EspSF) e o desempenho acadêmico. O estudo tem caráter observacional transversal analítico. As análises foram realizadas a partir de dados do curso de EspSF, que é ofertado na modalidade a distância, para médicos, dentistas e enfermeiros. Um total de 4438 alunos realizaram matrícula no curso até 2018, sendo considerados nas análises todos os concluintes até 2018 (n=4219). Para caracterização do desempenho acadêmico no curso foram considerados o estado do aluno ao final do curso, sendo os alunos classificados em: aprovado, reprovado, evadido ou desligado; e a nota final do aluno na atividade de casos complexos. Após uma análise descritiva foram construídos modelos de regressão logística para verificar a associação entre as características dos alunos e o desempenho no curso. A maior parte dos alunos é do gênero feminino (63,7%), brasileiro (63,9%), não é casado (66,9%), faz parte do Programa Mais Médicos (54,5%), formado há menos de um ano (32,0%), nunca havia realizado atividades de educação a distância (72,5%), tem acesso à internet em casa (94,9%) e no trabalho (64,5%), e usa o computador ao menos seis dias da semana (53,4%). Dentre os alunos há 2643 (52,6%) aprovados e 1576 (47,4%) reprovaram ou não concluíram o curso. Gênero, forma de entrada no curso, curso de graduação, idade no início do curso e a região de atuação profissional estão associados ao aproveitamento no curso e nota na avaliação de casos complexos. Palavras-chave: Desempenho acadêmico. Educação a distância. Ensino online. Atenção primária à saúde.

In this chapter, publications in the scientific literature that demonstrate the use of IoT in the context of medical decision-making were analyzed to present a review of recent developments and applications of the Internet of Things (IoT) for an enhanced medical decision. The main findings in this work have been to identify how intelligent systems can help improve the medical decision-making process, making an expanded summary of the area and pointing out challenges for the future of IoT and intelligent systems. Mapping was carried out in the scientific literature’s primary databases, identifying systematic review articles to support the chapter’s development. In total, fifty articles were analyzed, of which 28 are recently published systematic reviews, and 22 are secondary literature articles. The main conclusion has been about the areas of most extraordinary evidence for medical decisions and pointing out relevant works that have been developed in recent years, highlighting both the diagnostic of serious diseases and hospital management. In addition, with the emerging of a large volume of data, the challenge will be to model knowledge and apply reasoning techniques to recommend decisions in diagnoses of mild diseases, therapies, and even disease prevention.

Cancer is a group of diseases caused by the abnormal and disorderly growth of cells, representing the second leading cause of deaths worldwide. The number of cancer cases is growing yearly, medical systems are an essential tool to speed up the diagnosis process and increase patient survival probabilities. Electronic health record systems store the patient’s health data, which can be of structured and unstructured types. Physicians use all the information available in these systems during a cancer diagnostic, regardless of its sort and modality. Deep learning is a machine learning sub-field that has algorithms able to process data end-to-end using deep architectures, often inspired by the brain’s synaptic model. Using more than one data source in these architectures is known as multimodal deep learning. Computer-aided detection and diagnosis systems developed using multimodal deep learning algorithms have been achieving promising results in diagnostic performance, which are often comparable to human specialists. E-health and telemedicine applications can be boosted with these high-performance detection and diagnosis systems, which can provide real-time analysis capabilities to the healthcare staff and improve the quality of the medical services. This chapter explores the theory and the applications of multimodal deep learning techniques to develop computer-aided detection and diagnosis systems for cancer, describing examples of the state-of-the-art in this area, looking for essential and innovative aspects of these systems. Heterogeneous and hybrid fusion strategies combining high-quality imaging, clinical attributes and genetic data in deep multimodal architectures are resulting in computer-aided systems for cancer with promising diagnosis performance.

RecOAComp é um sistema para a recomendação de objetos de aprendizagem (OAs) baseado no perfil de competências do sujeito. A versão aplicativo (App), foi desenvolvido com o objetivo de atender os usuários que utilizam o sistema no smartphone. Assim, a versão App contempla especificidades relacionadas à experiência deste sujeito e tem potencial para colaborar na indicação de OAs, os quais possam ajudar o usuário na construção de um ou mais elementos das competências.

Os repositórios educacionais proporcionam oportunidade de acesso a objetos de aprendizagem elaborados por educadores de diferentes regiões e culturas. No entanto quando estudantes utilizam recursos educacionais não adaptados às suas características culturais, eles podem apresentar conflitos no seu processo de aprendizagem desestimulando o seu envolvimento. Nesse contexto, este artigo apresenta um modelo de sistema de recomendação educacional que se baseia nas dimensões culturais da aprendizagem para produzir recomendações de objetos culturalmente adaptados ao perfil educacional dos estudantes.

Background It is necessary to ensure functional diagnosis and auditory rehabilitation as part of a continuous and inseparable follow-up process that begins with Neonatal Hearing Screening to achieve the expected outcome in children with hearing loss. Different software controls the data of this process, adopting different strategies and involving the technology for this. However, there is no specific model available in the literature for analyzing the quality of the software aimed at recording and monitoring data from Neonatal Hearing Screening. Objective To propose a specific model for the analysis of the quality of softwares used for monitoring Neonatal Hearing Screening data, based on the ISO/IEC 25,010/2011 standards. Methods This is an applied research, in which a model was proposed, applied and evaluated to analyze the quality of Neonatal Hearing Screening softwares, based on an exploratory documental analysis of softwares related to the Neonatal Hearing Screening record domain. The quality model was proposed based on the ISO/IEC 25,010/2011 standards, constituting a checklist for qualitative analysis. Results For the Neonatal Hearing Screening registration softwares quality evaluation model, general criteria were presented, classified into eight main categories: Functional Adequacy, Performance Efficiency, Compatibility, Usability, Reliability, Safety, Maintenance and Portability. Specific criteria were presented in question format. Finally, a checklist for quality control and decision making was proposed. This model was evaluated by specialists and was approved. Conclusion The quality model presented in this article introduced important general and specific criteria to analyze softwares for Neonatal Hearing Screening. This model has been validated by specialists in informatics and audiology. Therefore, this model can be used comprehensively, as a standard assessment tool for Neonatal Hearing Screening softwares, allowing predicting improvements. It is suggested that the audiologists and informatics responsible for softwares of this nature, consider, in each country, the socioeconomic and health context to validate its applicability.

Objetiva-se descrever as consequências no aleitamento e na alimentação que terão as crianças órfãs menores de cinco anos em decorrência da morte materna, aplicando-se softwares livres de mineração de texto. Estudo transversal com base em artigos publicados nos repositórios PubMed e BIREME nos temas de morte materna e crianças órfãs. Foram selecionados dez artigos publicados entre 2005 e 2015, de acesso livre, nos quais foram lidos apenas o título ou o resumo e que cumpriam com os critérios. Os arquivos de texto definiram o corpus para análise de conteúdo semiestruturado. Palavras-chave foram incluídas para a mineração. A análise do corpus foi feita com TagCrowd e Textalyser para encontrar os termos mais e menos frequentes, AntConc e Voyant Tools, para extrair palavras-chave na análise de contexto. Foram analisadas 67.642 palavras em dez textos semiestruturados. Os termos CHILDREN (827) e DEATH (821) foram os mais frequentes, e os menos frequentes foram BREASTFEEDING (10) e NUTRITION (4). Foram encontradas 44 concordâncias para o termo raiz BREAST* e 25 para a palavra NUTRITION em orações como: “crianças órfãs têm o aumento de risco de mortalidade por falta de amamentação, e são mais susceptíveis às infecções”. As sentenças de concordância apontam que a mudança no aleitamento materno conduz a uma nutrição pobre, o que deixa o recém-nascido exposto a infecções, aumentando o risco de morte. O processamento de texto com as ferramentas livres foi rápido e permitiu extrair informações úteis e compreensíveis; a análise dos dez artigos mostrou as consequências na alimentação da criança após a morte materna, tendo efeito na morbidade e mortalidade infantil. The study aims to describe the consequences for future feeding of orphans under five years of age, following the mother’s death, applying open-access text mining software packages. This was a crosscutting study of articles indexed in PubMed and BIREME on the themes of maternal death and orphan children. We selected ten open-access articles published from 2005 to 2015 in which only the title or abstract were read and which met the selection criteria. The text files determined the corpus for analysis of the semi-structured content. Keywords were included for the mining. Analysis of the corpus used TagCrowd and Textalyser to analyze the more versus less frequent terms, and AntConc and Voyant Tools were used to extract keywords in the context analysis. A total of 67,642 words were analyzed in ten semi-structured texts. The terms CHILDREN (827) and DEATH (821) were the most frequent, while the least frequent were BREASTFEEDING (10) and NUTRITION (4). We found 44 concordances for the root term BREAST* and 25 for the word NUTRITION in phrases such as “orphan infants have increased risk of mortality due to lack of breastfeeding and are more susceptible to infections”. The concordant sentences indicate that the change away from maternal breastfeeding leads to poor nutrition and leaves the child more susceptible to infections, increasing the risk of death. Text processing with open-access tools was rapid and allowed extracting useful and comprehensible information; analysis of the ten articles showed the consequences for the child’s feeding after the mother’s death, affecting infant morbidity and mortality. El objetivo del trabajo fue describir las consecuencias en la lactancia y la alimentación que tendrán los niños huérfanos menores de cinco años, como consecuencia de la muerte materna, aplicando softwares libres de minería de textos. Se trata de un estudio transversal, a partir de los artículos publicados en los repositorios PubMed y BIREME, sobre los temas muerte materna y niños huérfanos. Se seleccionaron diez artículos, publicados entre 2005-2015, de acceso libre en los que se había leído solamente el título o el resumen y que cumplían con los criterios. Los archivos de texto definieron el corpus para el análisis de contenido semiestructurado. Se incluyeron palabras claves para la minería. El análisis del corpus se realizó con TagCrowd y Textalyser para el análisis de los términos más y menos frecuentes, AntConc y Voyant Tools, para extraer palabras-clave en el análisis de contexto. Se analizaron 67.642 palabras en diez textos semiestructurados. Los términos CHILDREN (827) y DEATH (821) fueron los más frecuentes, mientras que los menos frecuentes fueron BREASTFEEDING (10) y NUTRITION (4). Se encontraron 44 concordancias para el término raíz BREAST* y 25 para la palabra NUTRITION en oraciones como: “niños huérfanos tienen aumento de riesgo de mortalidad por ausencia de la lactancia, siendo más susceptibles a las infecciones”. Las frases de concordancia indican que el cambio en la lactancia materna conduce a una nutrición pobre, dejando al recién nacido expuesto a infecciones e incrementándose el riesgo de muerte. El procesamiento de texto con las herramientas libres fue rápido y permitió extraer información útil y comprensible; el análisis de los diez artículos mostró las consecuencias en la alimentación del niño tras la muerte materna, teniendo un efecto en la morbilidad y mortalidad infantil.

In the last few decades, obesity has grown exponentially and it progression is imminent contributing to the increase of mortality levels. Artificial Intelligence (AI), which is Computer Science area, could be well applied to management of obesity, such as an important tool to avoid the threat caused by this disease. The aim of this literature review was to show AI applications to obesity management and discuss their effectiveness. The search was performed in the following databases: Public Medline (PubMed), Web of Science, Biblioteca Regional de Medicina (BIREME) and Google Academic, by using following keywords, “artificial intelligence” and “obesity”. Our results led to some Artificial Intelligence systems used in obesity handling, which were: the Decision Support System to bariatric surgery patients; the MOPET app to motivate physical activity; Parameter Decreasing Methods and Artificial Neural Network to correlate obesity to cardiovascular disease; Artificial Neural Network to predict resting energy expenditure; a Neuro-Fuzzy Model to refine body mass index result; an Image Processing Algorithm; and a Support Vector Machine that monitors food intake. In this review, all investigated AI systems may have a tendency to more accurate results indicating a promising tool to manage obesity and related diseases.

Background: Continuing education of healthcare workers (HCWs) is an essential strategy for the control of tuberculosis (TB) transmission, enabling HCWs in early detection and appropriate treatment of TB cases. Methods: We developed a distance learning (DL) course on TB for nurses. We conducted a quasi-experimental before and after study to evaluate the DL community at the participant's learning level. In addition, to evaluate the DL community at the level of participant satisfaction, a cross-sectional study was carried out after the course. Nurses involved in active inpatient or outpatient care of patients were recruited to participate in the study. Results: Sixty-six participants started and completed the course and they were included in the analysis. The overall mean pretest and post-test scores were 10.3 ± 2.2 and 11.4 ± 2.7, respectively. Participants increased their knowledge to a statistically significant degree (p < 0.0001). At baseline, the frequency of correct answers was very low in some questions: number of people infected by Mycobacterium tuberculosis in the world (10.6%); number of TB cases in Brazil (36.4%); contagiousness of latent TB infection (LTBI) (28.8%); and definition of active case finding (45.5%). Course feedback was mostly positive, with majority of users saying they were satisfied or totally satisfied. Conclusions: A brief DL course on TB was associated with some improvement in knowledge among nurses. The baseline knowledge was low regarding TB epidemiologic data, concepts on LTBI, and active case finding. This finding emphasizes the need to further improve the competencies and knowledge of nurses.

Background According to UNAIDS, an estimated 100,000 cases of new HIV infections were reported in Latin America in 2017. At the time, Antiretroviral Therapy (ART) was already being offered for these new cases, and was largely prescribed in the Primary Health Care (PHC) system. The objective of this study is to evaluate the epidemiological profile of new cases of HIV infection in a Porto Alegre City, with some of the worst HIV indicators in Latin America. Methods A cross-sectional study, was carried out using a time series analysis, with data from Brazilian Information Systems. Data mining with WEKA software had been applied to evaluate the variables. The inclusion criteria were HIV diagnosis in the 2017 year, over 13 years who live in Porto Alegre City. Results The incidence of HIV cases was 777 in Porto Alegre, in which the population is 1.500.000 inhabitants. There were 308(39.6%) women, 71(34.8%) blacks and mulattoes and 40(5.1%) were pregnant. Of these, 60(7.7%) presented viral load above 100,000 copies/mL and 267(34.3%) below 10,000 copies/mL. The number of people with CD4 count below 200 was 26(3.3%) and 455(58.5%) people had a time interval shorter than 3 months between the diagnosis and the beginning of ART. ARTs with Tenofovir/Lamivudine/Dolutegravir and Tenofovir/Lamivudine/Efavirenz were prescribed in 566(95%) of the 590 patients who started ART. Among the 777 cases, 521(67%) were diagnosed through the Rapid Test, which are offered primarily in PHC. Conclusion The role of PHC was seen to be effective in reaching 90–90–90 Targets, since it has increased the diagnosis of HIV to immunocompetent individuals, facilitating easy access to first-line ART in short period of time. Future studies should be designed to investigate the role of PHC in the cascade of care for individuals with HIV. Information obtained from data mining can be applied to organize strategic interventions focused on local realities. Disclosure No significant relationships.

Background Sex Workers (SW) present greater vulnerability to sexually transmitted infections. There is a need to stimulate an active screening in the vulnerable population in the area covered by a Primary Health Care (PHC) team. The main goal of this study is to present an alternative healthcare assistance approach, applied to a specific area where a high index of SW can be found within the Southern Region of Brazil. Methods Experience case from a Brazilian PHU regarding adhesion to the unit, local interventions, and rapid test triage in SW was carried out in the health post, in order to promote sexual and reproductive healthcare in an interval of 5 years. Results Work related to the depicted area started with weekly visits as an approach strategy while offering health care services. Later on, a family planning was conducted and it included provinding access to condom. The SW didn’t accept the activities proposed by the PHC. The acceptance of the professional team amongst the community was slow, there was mistrust on the real interest for SW healthcare. The continuity of care was regarded as one os the strategies that was more effective to improve the quality of life of SW. Adhesion to prenatal services, offer of family planning and application of rapid testing for STIs were the most relevant activities that had been implemented. Within 173 applied tests, 135(78%) were women, of these 27(15,6%) presented pregnancy, and 11(6,3%) were trans women. 38(21,9%) were men. Within these tests, 10(6,3%) were positive for HIV, 35(20,3%) for syphilis and 1(0,6%) for HCV. Conclusion An effective intervention was possible when the health professional had the following scope: action at local level, engagement of local commerce, improve housing, and to involve the whole community in promoting better healthcare. Future studies should be designed to investigate the role of PHC in the SW healthcare. Disclosure No significant relationships.

Background Social engagement used as a tool for providing access to information and testing for Sexually Transmitted Infections (STIs) during events can be a promising strategy for reaching different population groups in public settings. The use of Rapid Tests (RT) for syphilis offers an unique opportunity to facilitate care in response to the syphilis epidemic. The objective of this study is to evaluate the epidemiological profile of people with positive RT for syphilis observed during RT events in Porto Alegre City, Brazil. Methods Cross-sectional study using a time series approach in all RTs carried out in Mobile unit RT events in 2018, through data mining, with WEKA software. Results The events were held in public spaces, such as squares and parks, included roundtable discussions, presentations, condom distribution, RTs for STIs, individual medical care with possibility of prescription and application of medication. Among 4156 RTs conducted, 626(15%) were positive for syphilis. The characteristics of the population tested were: 2003(48%) women, 374(9%) homo/bisexual, 710(17%) under 25 years and 1001(24%) elderly. In the subgroup with positive syphilis RTs, 250(39.9%) were women, 29(4.9%) homo/bisexual, 44(7%) under 25 years, 43(6.8%) positive RT for HIV and 43(6.8%) for Hepatitis C. Although the RTs for syphilis did not confirm diagnosis, the framework in place to deal with positive tests provides same-day medical consultation to evaluate the need for immediate treatment at the event site and referral for follow-up or continued treatment at the Primary Healthcare level. Conclusion Social interaction with the public during RT using a mobile healthcare units in places intended for leisure sought to increase awareness and access to people with undiagnosed STIs. Ensuring that specific subgroups of the general population have access to testing and medical consultation were shown to be important points when applying point of care tests. These aspects of community testing should be evaluated in future research. Disclosure No significant relationships.

The Recommender Systems in the area of Education have been investigated in order to minimize the information overload for whom the students are subjected, and to improve the learning processes. Personality traits are also researched as a characteristic in order to personalize the students' teaching-learning process. The purpose of this article is to present a doctoral research which proposes the development model of an Educational Recommender System based on the context, by using characteristics of the students' personality traits to customize the recommendations of learning objects.

Background The UNAIDS 90–90–90 programme aims to increase healthcare access for asymptomatic HIV patients, ensuring their well-being and disrupting the transmission network of the virus. The city of Porto Alegre, is one of the cities with the worst HIV indicators in Latin America. One of the strategies currently being implemented to address this is to develop a public Rapid Testing (RT) campaign, with immediate management of HIV, Syphilis and Hepatitis C. The objective of this research is to evaluate the initial results of the strategy in 2018. Methods Cross-sectional study using the time series approach for all RTs conducted in public testing events in 2018, through data mining, with WEKA software. Results Among the 4157 RTs carried out, reactive test results were observed as follows: 97(2.3%) HIV, 626(15%) syphilis and 156(3.7%) for Hepatitis C. The characteristics of this population are: 2004(48%) women, 375(9%) homo/bisexual, 710(17%) under 25 years and 1001(24%) elderly. In the HIV reactive subgroup, 29(30%) women, 29(30%) homo/bisexual, 11(11%) under 25 years, 17(17%) elderly, 54(56%) syphilis co-infection and 14(14%) for Hepatitis C reactive test, it should be noted that these results differ in proportion to published municipal/national data. Conclusion The availability of RTs in public spaces facilitated early diagnosis in young and homosexual/bisexual populations. However, the campaign also observed participation from other populations such as the elderly and homeless, who are guaranteed by law priority access to health services. The amplification of healthcare access through RTs made early diagnosis possible in asymptomatic individuals. According to data mining, this strategy reached population groups which differ from those observed in Epidemiological Bulletins of previous years. The recommendation to also encourage the provision of ART to the elderly and homeless can be an effective strategy that should be evaluated in future research. Disclosure No significant relationships.

This research proposal is about the integration between Learning Analytics and Cognitive Computing in the development of resources to support the process of personalized learning in Higher Education and from the use of these resources verify the improvement of motivation and engagement of students in learning processes. It is a work in progress project in which is offered customized processes to students, as well as prescriptive analysis of data to teachers to support decision-making and more assertive design of learning experiences.

Background Academic Detailing (AD) is a scientific dissemination strategy that is characterized by the one-by-one orientation to improve the quality of care. It has been used frequently in some developed countries about HIV management. HIV is a public health concern in all spheres of health care and its guidelines need to be constantly updated. The objective is to present the development of the virtual strategies for improvement of quality of health of HIV care in Latin America community-based Primary Care (PC) practices. Methods A descriptive study of the development of a learning object was carried out to improve effectiveness of care of HIV by the PC physicians. The structure and elements of a traditional AD strategy were adapted by an educational outreach program. Results The development of a learning object led to the compilation of materials necessary for the professional practice of HIV treatment, including models for notification, prescription, segmental examinations and the development of a flow for decision making, including multidisciplinary aspects such as contraception, social benefits, control of opportunistic infections, and others. The instructional design was structured to facilitate navigation, and can be used as support material during consultations, and can be easily accessed by mobile or computers. Conclusion After the START and TEMPRANO studies trials, which are tools extensively used in the management of HIV, PC practices assume a pivotal part of the care of people living with HIV. The Guidellines’ priority agenda of updating includes a number of other chronic and acute illnesses, however, especially in developing countries, there is no sheltered schedule for the physician to undertake lifelong education that includes so many updates. Virtual AD has the potential to reach out a large number of physicians, even with difficult access or time, without losing the one-by-one training feature. Disclosure No significant relationships.

Liver cancer was the fourth most deadly cancer in 2018 worldwide. Among liver cancers, hepatocarcinoma is the most prevalent cancer type. Diagnostic protocols are complex and suggest variations based on the patient’s context and the use of multiple data modalities. This paper briefly describes the steps involved in the development of a hepatocarcinoma computer-aided diagnosis using a multimodal deep learning approach with imaging and tabular data fusion. Data acquisition, preprocessing steps, architectural design decisions and possible use cases for the described architecture are discussed based on the partial results achieved on this ongoing research.

O presente artigo apresenta uma abordagem de Mineração de Dados Educacionais com o objetivo de identificar padrões de comportamento relacionados aos fatores motivacionais e aos estados de ânimo nas interações de alunos, em um ambiente virtual de aprendizagem. Para investigar esses padrões, aplicou-se o método de clusterização e utilizou-se o algoritmo K-means. Como resultado, obteve-se que a (des)motivação do aluno está diretamente relacionada com seu estado de ânimo, podendo sofrer alterações tanto positivas como negativas mediante o grau expresso nos fatores motivacionais de confiança, esforço e independência. This paper presents an Educational Data Mining approach to identify behavioral patterns related to motivational factors and moods in student interactions in a virtual learning environment. To investigate these patterns, the clustering method was applied and the K-means algorithm was used. As a result, it was found that the (dis) motivation of the student is directly related to their mood, and may undergo both positive and negative changes through the degree expressed in the motivational factors of confidence, effort and independence.

Este artigo apresenta uma solução m-Health para apoio à educação em saúde constituída por um aplicativo para Android no dispositivo móvel cliente e uma aplicação em máquina servidora. Diferentes conceitos e teorias estão integrados visando aumentar a motivação do usuário, com foco na mudança de comportamento para hábitos mais saudáveis e melhoria da qualidade de vida. Uma avaliação da interface da aplicação foi realizada através de questionário baseado na ISO 9241-11 e ABNT ISO/IEC 25062:2011 junto a oito especialistas. Além disto, o questionário WHOQOLbref foi disponibilizado aos usuários via app, e feedbacks sobre conteúdos também são analisados e trazidos para discussão This article presents an m-Health solution to support health education consisting of an Android application on the mobile device client and an application on a server machine. Different concepts and theories are integrated to increase the motivation of the user, focusing on changing behavior to healthier habits and improving the quality of life. An evaluation of the application interface was carried out through a questionnaire based on ISO 9241-11 and ABNT ISO / IEC 25062: 2011 with eight specialists. In addition, the WHOQOL-bref questionnaire was made available to users via the app, and many feedbacks on content are also analyzed and brought to the discussion.

Decisions about precautions and isolations are performed by healthcare professionals based on manuals, through tables and texts, not always available at the point of care. For this reason, the use of mobile application software to aid in decision making during health care is increasing. This article describes m-health application on standard precautions and based on transmission routes development to support the decision of healthcare workers while providing patient care. It is an applied research of technological production, where it was carried out prototyping, content database elaboration, creation, development and system evaluation. m-health application, called Isolation App, provides information on precautions and isolation, personal protective equipment suitable for each type of precaution, hand hygiene guidelines and contains national and international links related to the subject. Application was evaluated by healthcare workers experts in infection control and prevention. Results suggest the prototype may support health professionals, facilitating decision making at the point of care.