Journal
IEEE LATIN AMERICA TRANSACTIONS
Volume 19, Issue 6, Pages 944-951Publisher
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/TLA.2021.9451239
Keywords
COVID-19; X-ray imaging; Biomedical imaging; Deep learning; Image resolution; Coronaviruses; Visualization; alexnet; x-ray chest; convol; COVID-19; Deep Learning; recognition
Ask authors/readers for more resources
The study utilized deep learning with chest X-ray images to accurately detect COVID-19 related biomarkers with impressive accuracy, sensitivity, and specificity. Fine-tunned AlexNet network showed promising results in binary chest X-ray recognition between healthy individuals and those with COVID-19 or pneumonia symptoms, demonstrating the potential of AI in early diagnosis of respiratory illnesses.
The COVID-19 is a new disease from the virus SARS-CoV-2, the infection can cause respiratory illness with symptoms such as cough, fever, and, in severe cases, pneumonia. Early diagnosis is crucial for the correct treatment to reduce as much as possible the stress in the healthcare system. The need for auxiliary diagnostic tools has increased as there are no accurate automated toolkits available. Application of advanced artificial intelligence (AI) techniques coupled with radiological imaging can be helpful for the accurate detection of this disease. In this study, we have applied learning transfer to a convolutional neural network known as AlexNet for binary chest X-ray recognition (COVID-19 vs Healthy). We have fine-tunned AlexNet for our specific problem. The first layer, which works with RGB images, is replaced for images in a single intensity (grayscale). 11,312 chest X-ray images from six public databases were used to train the network. Among them are samples of healthy people and samples that present the effect of pneumonia and COVID-19 diseases. The results prove that deep learning with chest X-ray images can extract significant biomarkers related to COVID-19, since the obtained accuracy, sensitivity and specificity were 96.5%, 98.0%, and 91.7%, respectively. ROC analysis and confusion matrices are used to validate the results of the fine-tunned AlexNet network
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available