Artificial Intelligence” A system for identifying different types of epidemics using X-rays radiation

Abstract

The COVID-19 epidemic has put many lives in jeopardy from the very start. In order to classify epidemics, this study used the visual geometry group network (VGGNet). Pulmonary tuberculosis, normal lung, pneumonia, and COVID-19 were the four categories used to assess the 12068 chest X-ray photos collected from the Kaggle website. Using the chest X-ray pictures, we were able to identify and categorize the aforementioned condition using the VGGNet architecture. The accuracy, specificity, and sensitivity are some of the metrics used to evaluate these classes' efficacy. With regard to the parameters that were measured, the sensitivity, specificity, and accuracy were 0.98, 0.96, and 0.97, correspondingly. By correctly identifying variations in patients' X-ray pictures, our method may distinguish between various illnesses. The findings shown that when it comes to epidemic diagnosis, the VGG16 model may outperform VGG19. A quicker diagnosis and better prognosis for patients are both made possible by the VGG16-based method. In comparison to computed tomography (CT) pictures, the suggested model based on chest X-ray images is more accurate, simpler, and cheaper, according to the results.