Automatic Detection of Coronavirus Disease (COVID-19) in X-ray and CT\n Images: A Machine Learning-Based Approach

2020/04/22 by Sara Hosseinzadeh Kassani, Kassani, Sara Hosseinzadeh, Peyman Hosseinzadeh Kassasni +7
Computer Science · Medicine · #AI in cancer detection #COVID-19 diagnosis using AI #Computer Vision and Pattern Recognition (cs.CV) #Digital Imaging for Blood Diseases #FOS: Computer and information sciences #FOS: Electrical engineering #Image and Video Processing (eess.IV) #Radiomics and Machine Learning in Medical Imaging #electronic engineering #information engineering

paper · pdf · doi:10.48550/arxiv.2004.10641

openalex publication_date 2020/04/22 · openalex created_date 2022/07/26 · openalex updated_date 2026/07/28

Abstract

The newly identified Coronavirus pneumonia, subsequently termed COVID-19, is\nhighly transmittable and pathogenic with no clinically approved antiviral drug\nor vaccine available for treatment. The most common symptoms of COVID-19 are\ndry cough, sore throat, and fever. Symptoms can progress to a severe form of\npneumonia with critical complications, including septic shock, pulmonary edema,\nacute respiratory distress syndrome and multi-organ failure. While medical\nimaging is not currently recommended in Canada for primary diagnosis of\nCOVID-19, computer-aided diagnosis systems could assist in the early detection\nof COVID-19 abnormalities and help to monitor the progression of the disease,\npotentially reduce mortality rates. In this study, we compare popular deep\nlearning-based feature extraction frameworks for automatic COVID-19\nclassification. To obtain the most accurate feature, which is an essential\ncomponent of learning, MobileNet, DenseNet, Xception, ResNet, InceptionV3,\nInceptionResNetV2, VGGNet, NASNet were chosen amongst a pool of deep\nconvolutional neural networks. The extracted features were then fed into\nseveral machine learning classifiers to classify subjects as either a case of\nCOVID-19 or a control. This approach avoided task-specific data pre-processing\nmethods to support a better generalization ability for unseen data. The\nperformance of the proposed method was validated on a publicly available\nCOVID-19 dataset of chest X-ray and CT images. The DenseNet121 feature\nextractor with Bagging tree classifier achieved the best performance with 99%\nclassification accuracy. The second-best learner was a hybrid of the a ResNet50\nfeature extractor trained by LightGBM with an accuracy of 98%.\n

Citations

• A Decision-Theoretic Generalization of On-Line Learning and an Application to Boosting
• ImageNet classification with deep convolutional neural networks
• The epidemiology and pathogenesis of coronavirus disease (COVID-19) outbreak
• Transmission potential and severity of COVID-19 in South Korea
• COVID-19 infection: Emergence, transmission, and characteristics of human coronaviruses
• The impact of the COVID-19 epidemic on the utilization of emergency dental services
• COVID-Net: a tailored deep convolutional neural network design for detection of COVID-19 cases from chest X-ray images

Related

• Fast and accurate detection of Covid-19-related pneumonia from chest X-ray images with novel deep learning model
• Classification of COVID-19 X-ray Images Using a Combination of Deep and Handcrafted Features
• Comparative Analysis of State-of-the-Art Deep Learning Models for Detecting COVID-19 Lung Infection from Chest X-Ray Images
• Automatic Detection of COVID-19 and Pneumonia from Chest X-Ray using Deep Learning
• COVIDX-Net: A Framework of Deep Learning Classifiers to Diagnose COVID-19 in X-Ray Images