COVID-19 pneumonia detection using chest X-ray processing

Abstract

COVID-19 is one of the most fatal diseases and get increases rapidly through close contact with somebody infected or holding virus-contaminated objects. Another infectious illness referred to as Pneumonia is usually caused by infection because of a bacterium within the alveoli of the lungs. When contaminated tissue of the lungs has inflammation. To see if the patient has these diseases, experts conduct physical exams and identify their valetudinarian through Chest X-ray, ultrasound, or biopsy of the lungs. Misdiagnosis, inaccurate treatment, and ignored disease will cause the patient’s loss of life. Due to increasingly frequent cases, the number of COVID-19 test kits available in hospitals is decreasing. To prevent COVID-19 from spreading among humans, an autonomous detection method should be established as a quick, alternative diagnostic. For this purpose, three different deep learning models: Xception, VGG16, and Basic CNN have been proposed in this analysis for the diagnosis of patients infected with coronavirus, and pneumonia using an X-rays dataset of 6432 images is applied to prove the execution of the model. Xception model has the best efficiency among the two other models (VGG16 and Basic CNN). In terms of predicting cases, however, the Xception network outperformed the other models.