Search

First outbreak of COVID-19 was in the city of Wuhan in China in Dec.2019 and then it becomes a pandemic disease all around the world. World Health Organization (WHO) confirmed more than 5.5 million cases and 341,155 deaths from the disease till the time of writing this paper. This new worldwide disease forced researchers to make more precise way to diagnose COVID-19. In the last decade, medical imaging techniques show its efficiency in helping radiologists to detect and diagnose the diseases. Deep learning and transfer learning algorithms are good techniques to detect disease from different image source types such as X-Ray and CT scan images. In this work we used a deep learning technique based on Convolution Neural Network (CNN) to detect and diagnose COVID-19 disease using Chest X-ray images. Moreover, the modified AlexNet architecture is proposed in different scenarios were differing from each other in terms of the type of the pooling layers and/or the number of the neurons that have used in the second fully connected layer. The used chest X-ray images are gathered from two COVID-19 X-ray image datasets and one dataset includes large number of normal and pneumonia X-ray images. With the proposed models we obtained the same or even better result than the original AlexNet with having a smaller number of neurons in the second fully connected layer.

CNN-based transfer learning method plays a significant role in the detection of various objects such as cars, dogs, motorcycles, face and human detection in nighttime images by using visible light camera sensors. This method mainly depends on the images captured by cameras in order to detect the mentioned objects in a variety of environments based on convolutional neural networks (CNNs). In this study, we utilized the same method to detect coronavirus phenomena by using chest X-ray images that have been collected from three different open-source datasets with the aim of rapid detection of the infected patients and speed up the diagnostic process. We used one of the deep learning architectures in a Transfer Learning mode and modified its final layers to adapt to the number of classes in our investigation. The deep learning architecture that we used for the purpose of COVID-19 detection from X-ray images is a CNN designed to detect human in nighttime. We also modified the CNN architecture in three different scenarios named (Model 1, Model 2 and Model 3) in order to improve the classification results. Compared to model one and two, the result improved in model three and the number of misclassified cases reduced particularly in detecting Abnormal and COVID-19 cases. Although our CNN-based method shows high performance in COVID-19 detection, CNN decisions should not to be taken into consideration until clinical tests confirms symptoms of the infected patients.

The global spread of the COVID-19 is a continuously evolving situation and it is still a major risk on the health of people around the world. A huge number of people are infected by this deadly virus and the number is still getting increased day by day. At this time, no specific vaccines or treatments of COVID-19 are found. Numerous ways are offered to detect COVID-19 such as swab test, CDC and RT-PCR tests. All of them can detect corona virus in different ways but they are not recommended by the reason of their limited availability, inaccurate results, high false-negative rate predicates, high cost and time consuming. Hence, medical radiography and Computer Tomography (CT) images were suggested as the next best alternative of RT -PCR and other tests for detecting Covid-19 cases. Recent studies found that patients with COVID-19 cases are present abnormalities in chest X-Ray images. Motivated by this, many researchers propose deep learning systems for COVID-19 detection. Although, these developed AI systems have shown quite promising results in terms of accuracy, they are closed source and unavailable to the research community. Therefore, in the present work, we introduced a deep convolutional neural network design (SAARSNet) designed to detect COVID-19 cases from chest X-Ray images. 1292 X-Ray images have been used to train and test the proposed model. the images have been collected from two open-source datasets. The input images are progressively resized into (220 by 150 by 3) in order to decrease the training time of the system and improve the performance of the SAARSNet architecture. Furthermore, we also investigate how SAARSNet makes predictions under three different scenarios with the aim of distinguishing COVID-19 class from both Normal and Abnormal classes as well as gaining deeper perceptions into critical factors related to COVID-19 cases. We also used the confusion metrics for evaluating the performance of SAARSNet CNN in an attempt to measure the true and false identifications of the classes from the tested images. With the proposed architecture promising results has been achieved in all of the three different scenarios. Although, there are some misclassified cases of COVID-19, the corresponding performance was best in detecting both Normal and Abnormal cases correctly. Furthermore, in the three classes scenario, normal class has been achieved 100% positive predictive value while optimistic results have been investigated in detecting COVID-19 and abnormal classes.