Background: Although the world has been facing the COVID-19 pandemic for over a year, we understand that there are still some challenges in using Internet of Things (IoT) devices as allies in this fight. Among the main difficulties, we can mention the selection of appropriate devices and the correct measurement and subsequent analysis of previously obtained vital signs.
Methods: In this context, we present a condensed compilation of IoT devices to monitor the vital signs often used to monitor COVID-19. We focus on easy-to-use devices currently available on the market to the general user. Also, the presented analysis is helpful for long COVID-19 monitoring, which is particularly useful to governments and hospitals to analyze eventual sequels on those citizens who tested positive beforehand.
Results: The review resulted in 148 heterogeneous devices offering different capabilities. Our first contribution resides in detailing several aspects of each IoT device, indicating which are the most suitable for particular use-case situations. Moreover, our article introduces some challenges and insights into assembling a smart city composed of IoT devices.
Conclusion: Here, technological trends such as Serverless computing, homomorphic cryptography, Federated Learning, Elixir programming language, Web Assembly, and vertical elasticity are discussed towards enabling vital sign-driven data capturing and processing. Although there are several IoT devices for health monitoring, there is still work to standardize data formats and APIs for data extraction.
A técnica de aprendizado federado é muito utilizada quando os dados a serem usados pelos modelos de aprendizado de máquina são sensíveis ou sigilosos. No entanto, aprendizado federado prevê o treinamento dos modelos por conta do usuário, que nem sempre está disponível para treinar o modelo ou não possui recursos computacionais eficientes. Esse trabalho apresenta uma arquitetura para execução de aprendizado federado de maneira segura e eficiente utilizando os recursos de borda em hospitais inteligentes.
Glaucoma é a principal causa mundial de perda irreversível de visão. Afim de viabilizar a implantação de uma ferramenta de diagnóstico de glaucoma para a clínica médica, um trabalho base foi selecionado e otimizado. Ao unificar duas redes de segmentação reduzimos o tempo de processamento em 24,24%, e adicionando uma segunda rede de classificação direta aumentamos a sensitividade do modelo em 3%, em comparação com o trabalho base.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.