Dissertação de Mestrado de Noemi Maritza Lapa Romero

DEFESA DE DISSERTAÇÃO DE MESTRADO

Aluna: Noemi Maritza Lapa Romero
Orientador: Prof. Dr. João Luiz Dihl Comba

Título: COVID-VR: A Deep Learning COVID-19 Classification Model Using Volume-Rendered Computer Tomography
Linha de Pesquisa: Computação Gráfica e Visualização de Dados

Data: 29/06/2022
Horário: 10h
Esta banca ocorrerá excepcionalmente de forma totalmente remota. Interessados em assistir a defesa poderão acessar a sala virtual através do link https://us02web.zoom.us/j/85111446704?pwd=RWZqMHVHWkVMcTN4NlVSclQ1ZzVWdz09

Banca Examinadora:
– Prof. Dr. Rodrigo Coelho Barros (PUCRS)
– Prof. Dr. Claudio Rosito Jung (UFRGS)
– Prof. Dr. Thiago Lopes Trugillo da Silveira (UFRGS)
Presidente da Banca: Prof. Dr. João Luiz Dihl Comba

Abstract: The COVID-19 pandemic brought several challenges to health systems worldwide. Since most patients with COVID-19 have lung infections, a Computer Tomography (CT) of the chest was often used to identify COVID-19 infections, as well as other classes of pulmonary diseases. Deep-learning architectures surfaced to automatically identify classes of pulmonary diseases, using the slices of CTs as inputs to classification models. This work proposes COVID-VR, a novel approach for classifying COVID-19 based on volume rendering images of the lungs taken from different angles, thus providing a global view of the entire lung in each image. We compared our proposal against leading competing strategies with available solutions, using private data from partner hospitals and publicly available data. Results show that our approach successfully identifies pulmonary lesions and is competitive against slice-based methods. Although our experiments were focused on COVID-19 data, our solution is extensible to other pulmonary diseases.

Keywords: Deep Learning. Classification Models. Computer Tomography. Volume Rendering. COVID-19