Analysis of Imbalanced Datasets in the Performance of Deep Learning Approaches for COVID-19 Screening from Chest X-ray Imaging: Impact of Sex and Age Factors
Use este enlace para citar
http://hdl.handle.net/2183/36467Coleccións
Metadatos
Mostrar o rexistro completo do ítemTítulo
Analysis of Imbalanced Datasets in the Performance of Deep Learning Approaches for COVID-19 Screening from Chest X-ray Imaging: Impact of Sex and Age FactorsAutor(es)
Data
2023-02-16Cita bibliográfica
L. Alvarez, J. D. Moura, L. Ramos, J. Novo, y M. Ortega, «Analysis of Imbalanced Datasets in the Performance of Deep Learning Approaches for COVID-19 Screening from Chest X-ray Imaging: Impact of Sex and Age Factors», Kalpa Publications in Computing, vol. 14, pp. 174-177. doi: 10.29007/v25g.
Resumo
[Absctract]: In this work, we analysed 11 imbalance scenarios with female and male COVID-19 patients present in different proportions for the sex analysis, and 6 scenarios where only one specific age range was used for training for the age factor. In each study, 3 different approaches for automatic COVID-19 screening were used: (I) Normal vs COVID-19, (II) Pneumonia vs COVID-19 and (III) Non-COVID-19 vs COVID-19.
The present study was validated using two representative public chest X-ray datasets, allowing a reliable analysis to support the clinical decision-making process.
The results for the sex-related analysis indicate this factor slightly affects the COVID- 19 deep learning-based systems, although the identified differences are not relevant enough to considerably worsen the system. Regarding the age-related analysis, this factor was observed to be influencing the system in a more consistent way than the sex factor, as it was present in all considered scenarios.
Palabras chave
CAD system
Chest X-ray
COVID-19
Deep learning
Chest X-ray
COVID-19
Deep learning
Descrición
Comunicación presentada al V Congreso XoveTIC, organizado por el Centro de Investigación en TIC da Universidade da Coruña (CITIC), tendrá lugar los días 5 y 6 de octubre de 2022
Versión do editor
ISSN
2515-1762