Segmentation of CT-Scan Images Using UNet Network for Patients Diagnosed with COVID-19

Resource type
Title
Segmentation of CT-Scan Images Using UNet Network for Patients Diagnosed with COVID-19
Abstract
The use of computational tools for medical image processing are promising tools to effectively detect COVID-19 as an alternative to expensive and time-consuming RT-PCR tests. For this specific task, CXR (Chest X-Ray) and CCT (Chest CT Scans) are the most common examinations to support diagnosis through radiology analysis. With these images, it is possible to support diagnosis and determine the disease’s severity stage. Computerized COVID-19 quantification and evaluation require an efficient segmentation process. Essential tasks for automatic segmentation tools are precisely identifying the lungs, lobes, bronchopulmonary segments, and infected regions or lesions. Segmented areas can provide handcrafted or self-learned diagnostic criteria for various applications. This Chapter presents different techniques applied for Chest CT Scans segmentation, considering the state of the art of UNet networks to segment COVID-19 CT scans and a segmentation experiment for network evaluation. Along 200 epochs, a dice coefficient of 0.83 was obtained.
Book Title
Computerized Systems for Diagnosis and Treatment of COVID-19
Place
Cham
Publisher
Springer International Publishing
Date
2023
Pages
29-44
Language
en
ISBN
978-3-031-30788-1
Accessed
10/10/23, 4:37 AM
Library Catalog
Springer Link
Citation
Bernardo Gois, F. N., & Lobo Marques, J. A. (2023). Segmentation of CT-Scan Images Using UNet Network for Patients Diagnosed with COVID-19. In J. A. Lobo Marques & S. J. Fong (Eds.), Computerized Systems for Diagnosis and Treatment of COVID-19 (pp. 29–44). Springer International Publishing. https://doi.org/10.1007/978-3-031-30788-1_3
Cooperation