Classification of COVID-19 in chest X-ray images using DeTraC deep convolutional neural network

Asmaa Abbas; Mohammed Abdelsamea; Mohamed Gaber

Diagnosis of COVID-19 is typically associated with both the symptoms of pneumonia 2 and Chest X-ray tests. Chest X-ray is the first imaging technique that plays an 3 important role in the diagnosis of COVID-19 disease. Fig. 1 shows a negative example 4 of a normal chest x-ray, a positive one with COVID-19, and a positive one with the 5 severe acute respiratory syndrome (SARS). 6 In the last few months, World Health Organization (WHO) has declared that a new 7 virus called COVID-19 has been spread aggressively in several countries around the 8 world [1] . Fast detection of the COVID-19 can be contributed to control the spread of 9 the disease. One of the most successful algorithms that have been proved its ability to 10 diagnosis medical images with high accuracy is convolution neural network (CNN ). For 11 example, in [2] , a CNN was applied based on Inception network to detect COVID-19 level features. Those features were fed into a Support Vector Machine SVM as a 17 machine learning classifier in order to detect the COVID-19 cases. Moreover, in [5] , a 18 CNN architecture called COVID-Net based on transfer learning was applied to classify 19 the CXR images into four classes: normal, bacterial infection, non-COVID and 20 COVID-19 viral infection. 21 Several classical machine learning approaches have been previously used for 22 automatic classification of digitised chest images [6, 7] . For instance, in [8] , three 23 statistical features were calculated from lung texture to discriminate between malignant 24 and benign lung nodules using a support vector machine classifier. A grey-level 25 co-occurrence matrix method was used with Backpropagation Network [9] to classify 26 images from being normal or cancerous. With the availability of enough annotated 27 images, deep learning approaches [10, 11] have demonstrated their superiority over the 28 classical machine learning approaches. CNN architecture is one of the most popular 29 deep learning approaches with superior achievements in the medical imaging domain [12] . 30 The primary success of CNN is due to its ability to learn features automatically from 31 domain-specific images, unlike the classical machine learning methods. The popular 32 strategy for training CNN architecture is to transfer learned knowledge from a 33 pre-trained network that fulfilled one task into a new task [13] . This method is faster 34 and easy to apply without the need for a huge annotated dataset for training; therefore 35 many researchers tend to apply this strategy especially with medical imaging.