COVID-19 Diagnosis Systems Based on Deep Convolutional Neural Networks Techniques: A Review

Abstract

The rapidly spreading of the viral disease 'COVID-19' causes millions of infections and deaths worldwide. It causes a devastating impact on the lifestyle, public health, and the global economy. This motivates the researchers to invent and develop innovative and automated methods to detect COVID-19 at its early stages. It is necessary to isolate the positive cases quickly to prevent this epidemic and treat affected patients. Many diagnosis methods are proposed to perform accurate and fast detection for COVID-19, such as Reverse Transcription-Polymerase Chain Reaction (RT -PCR). The clinical studies indicate that the severity of COVID-19 cases depends on the virus's amount within infected lungs. Chest X-ray (CXR) and Computed Tomography (CT) images are useful imaging methods for diagnosing COVID-19 cases. Deep Convolutional Neural Network (DCNN) is a machine learning technique usually used in computer vision applications. This review focuses on utilizing the DCNN methods for building an automated Computer-Aided Diagnosis (CADs) system to detect and classify the infected cases of the COVID-19 disease accurately and fast. These techniques are used to extracts valuable information by analyzing a massive amount of CXR and CT images that can critically impact on screening of Covid-19. DCNN techniques proved their robustness, potentiality, and advancement by comparing them among the other learning algorithms. It is worth noting that DCNN is an essential tool for supporting the physicians' clinical decisions.