Diagnosis and Localization of COVID-19 Using Deep Transfer Learning and Model Weighting

In this paper, we present a novel approach to diagnose and localize COVID-19 using CT and X-ray images, which is based on deep transfer learning. The proposed method utilizes a pre-trained convolutional neural network (CNN) architecture and fine-tunes it on a decent-sized COVID-19 dataset. In addition, we introduce a new model weighting technique to enhance the performance of the proposed method. The experimental results demonstrate that the proposed approach outperforms existing state-of-the-art methods in terms of accuracy, sensitivity, specificity, and area under the receiver operating characteristic curve (AUC-ROC) for both diagnosis and localization tasks. This method has the potential to aid radiologists and clinicians in accurately and quickly diagnosing COVID-19 in clinical settings.