Por favor, use este identificador para citar o enlazar este ítem: http://conacyt.repositorioinstitucional.mx/jspui/handle/1000/8890
Novel Deep CNNs Explore Regions, Boundaries, and Residual Learning for COVID-19 Infection Analysis in Lung CT
Dr. Saddam Hussain Khan
Acceso Abierto
Atribución-NoComercial-SinDerivadas
https://doi.org/10.1101/2024.05.08.24307035
https://www.medrxiv.org/content/10.1101/2024.05.08.24307035v2
Abstract COVID-19 poses a global health crisis, necessitating precise diagnostic methods for timely containment. However, accurately delineating COVID-19 affected regions in Lung CT scans is challenging due to contrast variations and significant texture diversity. In this regard, this study introduces a novel two-stage classification and segmentation CNN approach for COVID-19 lung radiological pattern analysis. A novel Residual-BRNet is developed to integrate boundary and regional operations with residual learning, capturing key COVID-19 radiological homogeneous regions, texture variations, and structural contrast patterns in the classification stage. Subsequently, infectious CT images undergo lesion segmentation in the second stage using the newly proposed RESeg segmentation CNN. The RESeg leverages both average and max-pooling implementations to simultaneously learn region homogeneity and boundary-related patterns. Furthermore, novel pixel attention (PA) blocks are integrated into RESeg to effectively address mildly infected regions. The evaluation of the proposed Residual-BRNet CNN demonstrates promising performance metrics, achieving an accuracy of 97.97%, F1-score of 98.01%, sensitivity of 98.42%, and MCC of 96.81%. Meanwhile, PA-RESeg achieves optimal segmentation performance with an IoU score of 98.43% and a Dice Similarity score of 95.96% of the lesion region. These findings highlight the potential of the proposed diagnosis framework to assist radiologists in identifying and analyzing COVID-19 affected lung regions. The CAD GUI diagnosis tool is provided at https://github.com/PRLAB21/COVID-19-Diagnostic-System.
bioRxiv
17-05-2024
Preimpreso
Inglés
Público en general
VIRUS RESPIRATORIOS
Aparece en las colecciones: Materiales de Consulta y Comunicados Técnicos

Cargar archivos: