| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 30.65 MB | Adobe PDF |
Authors
Advisor(s)
Abstract(s)
O papel da tecnologia na indústria da medicina tem promovido grandes avanços e trans formações profundas. Tornar os sistemas de saúde mais pro-ativos, modernos e eficazes
é o principal foco desta revolução tecnológica. O surgimento da pandemia COVID 19, não só acelerou, como impulsionou este crescimento. Principalmente, no que diz
respeito aos meios de diagnóstico, dado que, na área da saúde, quanto mais rápido se
diagnosticar, mais rápido se pode tratar. No sentido de melhorar as decisões estratégicas
através da obtenção de informação mais objetiva, foram criados modelos e propostas que
pretendem oferecer serviços mais autónomos e ao mesmo tempo seguros e ágeis, sem
comprometer a eficiência.
Esta dissertação tem como objetivo apresentar uma proposta de uma plataforma para
realizar auscultação autónoma nas costas das pessoas, capaz de ajudar o profissional de
saúde a realizar avaliações essenciais que contribuam para o bem-estar do utente.
A partir de informações externas do corpo obtidas através de um modelo de previsão,
foi possível calcular os pontos de auscultação. Estes pontos de auscultação serão depois
transformados em coordenadas espaciais e inseridos num braço robô colaborativo. Este
braço robótico é responsável por realizar o contacto entre a superfície do estetoscópio e
o corpo do paciente, nos vários pontos de auscultação desejados.
Esta ferramenta, ao ser utilizada nos locais adequados para o efeito, apresenta van tagens tais como, a automatização do processo de auscultação por parte do prestador
de cuidados, disponibilizando este para tarefas mais urgentes, e o possível rastreio de
doenças mais cedo.
Para avaliar a plataforma de auscultação autónoma desenvolvida, esta foi colocada
em prática nas costas de seis pacientes. Com estes testes foi possível concluir que o braço
robótico simula a colocação do estetoscópio nos pontos desejados de auscultação com
sucesso, dependendo da estrutura física do paciente. Apesar disso, devem ser aplicadas
melhorias nesta plataforma, principalmente, no que toca à estrutura da mesma.
he role of technology in the medical industry has promoted major advances and pro found transformations. Making healthcare systems more proactive, modern, and effec tive is the main focus of this technological revolution. The emergence of the COVID-19 pandemic has not only accelerated, but driven this growth. Especially with regard to di agnostic means, since in healthcare, the faster you can diagnose, the faster you can treat. In order to improve strategic decisions by obtaining more objective information, models and proposals have been created that intend to offer more autonomous and at the same time safer and more agile services, without compromising efficiency. This dissertation aims to present a proposal for a platform to perform autonomous auscultation on people’s backs, capable of helping the health professional to perform essential assessments that contribute to the user’s well-being. From external body information obtained through a prediction model, it was possi ble to calculate auscultation points. These auscultation points will then be transformed into spatial coordinates and inserted into a collaborative robot arm. This robotic arm is responsible for making contact between the stethoscope surface and the patient’s body at the various auscultation points desired. This tool, when used in the appropriate locations, has advantages such as automating the auscultation process for the caregiver, making the caregiver available for more urgent tasks, and possible earlier screening for diseases. To evaluate the developed autonomous auscultation platform, it was put into practice on the back of six patients. With these tests it was possible to conclude that the robotic arm simulates the placement of the stethoscope in the desired auscultation points succes sfully, depending on the physical structure of the patient. Nevertheless, improvements should be made to this platform, especially with regard to its structure.
he role of technology in the medical industry has promoted major advances and pro found transformations. Making healthcare systems more proactive, modern, and effec tive is the main focus of this technological revolution. The emergence of the COVID-19 pandemic has not only accelerated, but driven this growth. Especially with regard to di agnostic means, since in healthcare, the faster you can diagnose, the faster you can treat. In order to improve strategic decisions by obtaining more objective information, models and proposals have been created that intend to offer more autonomous and at the same time safer and more agile services, without compromising efficiency. This dissertation aims to present a proposal for a platform to perform autonomous auscultation on people’s backs, capable of helping the health professional to perform essential assessments that contribute to the user’s well-being. From external body information obtained through a prediction model, it was possi ble to calculate auscultation points. These auscultation points will then be transformed into spatial coordinates and inserted into a collaborative robot arm. This robotic arm is responsible for making contact between the stethoscope surface and the patient’s body at the various auscultation points desired. This tool, when used in the appropriate locations, has advantages such as automating the auscultation process for the caregiver, making the caregiver available for more urgent tasks, and possible earlier screening for diseases. To evaluate the developed autonomous auscultation platform, it was put into practice on the back of six patients. With these tests it was possible to conclude that the robotic arm simulates the placement of the stethoscope in the desired auscultation points succes sfully, depending on the physical structure of the patient. Nevertheless, improvements should be made to this platform, especially with regard to its structure.
Description
Keywords
Autonomous Auscultation Machine Learning MobileNet UNet InceptionResNet Adam Artificial Intelligence Robotics Healthcare Computer Vision