Application of DenseNet in Identifying Pneumonia in Chest X-Rays

• Roney Nogueira de Sousa IFCE
• Maria Elizabeth de Aguiar Lima IFCE
• Francisca Raquel de Vasconcelos Silveira IFCE

DOI: https://doi.org/10.5753/ercas.2024.238513

Abstract

We evaluated the effectiveness of DenseNet in pneumonia diagnosis using public chest X-ray data. Employing data augmentation techniques and training for 100 epochs, we achieved satisfactory results: accuracy of 95.67%, AUC of 98.52%, precision of 96.20%, and recall of 96.69%. These findings underscore the effectiveness of the architecture in classifying X-ray images for pneumonia diagnosis.

References

Chollet, F. et al. (2015). Keras. [link].

da Silva Bomfim, V. V. B., de Andrade Brandão, M., et al. (2023). Aspectos radiológicos no diagnóstico de pneumonia. Revista Ibero-Americana de Humanidades, Ciências e Educação, 9(5):2523–2532.

Gavrikov, P. (2020). Visualkeras. [link].

GBD 2015, C. R. D. C. et al. (2017). Global, regional, and national deaths, prevalence, disability-adjusted life years, and years lived with disability for chronic obstructive pulmonary disease and asthma, 1990–2015: a systematic analysis for the global burden of disease study 2015. The Lancet. Respiratory Medicine, 5(9):691.

Hunter, J. D. (2007). Matplotlib: A 2d graphics environment. Computing in Science & Engineering, 9(3):90–95.

Van Rossum, G. and Drake, F. L. (2009). Python 3 Reference Manual. CreateSpace, Scotts Valley, CA.