Requisitos de qualidade dos sinais para modelagem hemodinâmica em medicina intensiva: desafios e oportunidades
Arquivos
Data
2023-07-01
Autores
Almeida, Gustavo Henrique de [UNIFESP]
Orientadores
Aletti, Federico
Tipo
Trabalho de conclusão de curso de graduação
Título da Revista
ISSN da Revista
Título de Volume
Resumo
The proper monitoring of critically ill patients in Intensive Care Units (ICU) is essential to the decision-making of clinical teams being, in many cases, a key factor of patient’s outcome. In this sense, physi ological variables such as arterial pressure, heart rate and respiratory rate and O2 saturation are continuously monitored, generating large amounts of data that can be useful for the development of models to aid in the monitoring of those patients, providing alarm signals through the ICU monitors and even predicting the out come. However, some quality requisites are crucial to the reliability of the model, such as reduction of noise sources, technical specifications of the acquisition system, preprocessing with attenuation of artifacts, sampling and missing-data analysis, choice of the best features and so on. This study aims to propose a brief discussion about signal quality requisites to hemodynamic modelling in Intensive Medicine, and how the combination of data driven approaches (such as AI and big data algorithms) and hypothesis-driven approaches (such as physiological modelling) can help in feature selection and improve the modeling approaches so that they can indeed be applied to bedside monitoring