CoviDet: COVID-19 Case Detection Utilizing Deep Neural Networks Using X-Ray Images

Abstract

Many people developed serious respiratory issues as a result of the Coronavirus illness outbreak, which was recognized as a global health threat. Because the virus initially targets the lungs in the human body, chest X-ray imaging features were thought to be effective for early identification of the illness. The chest X-ray data of 141 infected patients and 142 non-infected patients (112 images in training dataset and 30 in test dataset each) was utilized to develop a CNN (Convolutional Neural Network) model for early disease identification in this study, which was based on open data from Cohen J., Morrison P., Dao L., 2020. The model was trained with 100 epochs of both infected and non-infected people's chest X-ray pictures, resulting in a final accuracy of 0.97. In order to use this model as a professional diagnosis element, it is highly recommended it be improved with more images and the model can be restructured to get a better accuracy.