Accurate Detection of COVID-19 and Pneumonia from Chest X-Rays and CT Images Using DNN
K. Sekar, Suganya Devi K, Murali Baskaran, T Dheepa
Abstract
X-rays play a crucial role in computed tomography (CT) imaging, enabling the visualization of internal structures within the body. It is particularly valuable in assessing pulmonary injuries in COVID-19 patients. Global health is facing a serious crisis as a result of such disease. Most affected persons could be saved if COVID-19 is diagnosed in time. However, COVID-19 can sometimes be misdiagnosed as other lung diseases, potentially leading to severe consequences such as death or significant injury, especially if the virus spreads rapidly throughout the thoracic cells. This research proposes a multi-classification deep neural network(DNN) to identify COVID-19 and pneumonia using a combination of CX-R and CT scan image datasets. The combined dataset will improve classification performance. The proposed DNN model achieves significant results in detecting COVID-19 and pneumonia. The proposed model provides an accuracy of 98.75%, a recall value of 98.75%, a precision value of 98.45%, <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$F_{1}$</tex> -score of 98.43%, and an AUC of 99.60% on the above datasets.