A Deep Learning Approach in Detection of COVID-19 Positive Patients using CT Scan Images

This paper presents a method for evaluating the utility of deep transfer learning in the development of a classifier for detecting COVID-19 positive patients using CT scan images. Deep Learning (DL) is good at detecting COVID-19 cases, according to the research. For expanding the training dataset to reduce overfitting and improve the model's generalization capacity data augmentation approach is employed. The proposed study has evaluated a set of pretrained deep neural networks for Convolutional Neural Network (CNN). The suggested model used DenseNet with Res Net and a two-layer CNN model which gives better performance. The proposed model gives efficient results with the training and testing accuracy of 0.98 and 0.96.