Empowering Diagnosis: A Review On Deep Learning Applications for COVID-19 and Pneumonia in X-Ray Images

Yaqoub, Karin Younis
Abdulazeez, Adnan Mohsin

JISA (Jurnal Informatika dan Sains)
Vol 7, No 1 (2024): JISA(Jurnal Informatika dan Sains)

The emergence of COVID-19, a highly contagious virus capable of infecting both the upper and lower respiratory tracts, has led to one of the deadliest pandemics in modern history, claiming millions of lives worldwide. Early and accurate detection of this rapidly spreading disease is crucial for effective containment and saving lives. Chest X-ray (CXR) stands out as a promising diagnostic tool due to its accessibility, affordability, and long-term sample preservation. However, distinguishing COVID-19 pneumonia from other respiratory ailments poses a significant challenge. This article delves into various approaches utilized for COVID-19 detection and the hurdles encountered in this endeavor. The imperative for developing automated detection systems to mitigate virus transmission via contact is underscored. Notably, deep learning architectures such as ResNet, Inception and Googlenet have been deployed for COVID-19 detection, albeit with a focus on identifying pneumonia cases. Discriminating between COVID-19-induced pneumonia and pneumonia caused by other pathogens remains a formidable task, demanding innovative solutions for accurate and timely diagnosis.