Abstract
Since the outbreak of Coronavirus disease 2019 (COVID-19), it has been spreading rapidly worldwide and has not yet been effectively controlled. Many researchers are studying novel Coronavirus pneumonia from chest X-ray images. In order to improve the detection accuracy, two modules sensitive to feature information, dual-path multiscale feature fusion module and dense depthwise separable convolution module, are proposed. Based on these two modules, a lightweight convolutional neural network model, D2-CovidNet, is designed to assist experts in diagnosing COVID-19 by identifying chest X-ray images. D2-CovidNet is tested on two public data sets, and its classification accuracy, precision, sensitivity, specificity, and F1-score are 94.56%, 95.14%, 94.02%, 96.61%, and 95.30%, respectively. Specifically, the precision, sensitivity, and specificity of the network for COVID-19 are 98.97%, 94.12%, and 99.84%, respectively. D2-CovidNet has fewer computation number and parameter number. Compared with other methods, D2-CovidNet can help diagnose COVID-19 more quickly and accurately.
Highlights
Coronavirus disease 2019 (COVID-19) caused by the 2019 novel Coronavirus (2019-nCoV) has spread all over the world in a very short time. e detection technology based on reverse transcription polymerase chain reaction (RTPCR) is the most widely used method for diagnosing COVID-19
Dataset. e chest X-ray images used in our experiments are from two open-source datasets. e common pneumonia and normal chest X-ray images are selected from the data set provided by Kaggle [26]
It contains a total of 5863 chest X-rays images. 4265 chest X-ray images of pneumonia and 1575 normal chest X-ray images are selected from this dataset. e COVID-19 chest X-ray images used in our experiments are selected from the dataset collected by Cohen et al [27]. is dataset contains 790 chest X-ray images and computed tomography (CT) images of infected patients with COVID-19 or other pneumonia. 412 chest X-ray images of with COVID-19 patients are selected from this dataset
Summary
Coronavirus disease 2019 (COVID-19) caused by the 2019 novel Coronavirus (2019-nCoV) has spread all over the world in a very short time. e detection technology based on reverse transcription polymerase chain reaction (RTPCR) is the most widely used method for diagnosing COVID-19. Wang et al [6] proposed and improved a deep learning approach with global average pooling (GAP) to classified colonoscopy polyp images for assisted diagnosis. Wang et al [7] designed the channel feature weight extraction module (CFWE) according to the characteristics of chest X-ray image and proposed a new CFW-Net. Apostolopoulos and Mpesiana [8] used a CNN based on the transfer learning method to automatically detect X-ray images. Khan et al [10] proposed a deep CNN model based on Xception-CoroNet and used a pretraining method based on the ImageNet dataset to identify COVID-19-positive chest X-ray images. Ozturk et al [11] proposed a DarkCovidNet model with fewer parameters to automatically detect COVID-19-positive chest X-ray images
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
