Classification of the COVID-19 infected patients using DenseNet201 based deep transfer learning

TLDR

Deep learning, especially transfer learning with pre‑trained networks, is widely used for radiological image analysis but its application to chest CT for COVID‑19 diagnosis remains limited. The study aims to develop an automated tool that classifies chest CT scans as COVID‑19 positive or negative using pre‑trained deep learning models. A DenseNet201‑based deep transfer learning model, leveraging ImageNet‑pretrained weights and a convolutional architecture, was trained and evaluated on chest CT images. The proposed DTL model outperformed competing approaches in classifying COVID‑19 from chest CT scans.

Abstract

Deep learning models are widely used in the automatic analysis of radiological images. These techniques can train the weights of networks on large datasets as well as fine tuning the weights of pre-trained networks on small datasets. Due to the small COVID-19 dataset available, the pre-trained neural networks can be used for diagnosis of coronavirus. However, these techniques applied on chest CT image is very limited till now. Hence, the main aim of this paper to use the pre-trained deep learning architectures as an automated tool to detection and diagnosis of COVID-19 in chest CT. A DenseNet201 based deep transfer learning (DTL) is proposed to classify the patients as COVID infected or not i.e. COVID-19 (+) or COVID (−). The proposed model is utilized to extract features by using its own learned weights on the ImageNet dataset along with a convolutional neural structure. Extensive experiments are performed to evaluate the performance of the propose DTL model on COVID-19 chest CT scan images. Comparative analyses reveal that the proposed DTL based COVID-19 classification model outperforms the competitive approaches.