An Enhanced Convolutional Neural Network for COVID-19 Detection

The recent novel coronavirus (COVID-19, as the World Health Organization has called it) proven to be a source of risk for global public health. virus, which causes an acute respiratory disease in persons, spreads rapidly and is now threatening more than 150 countries around world. One essential procedures that patients with COVID-19 need accurate rapid screening process. In this research, utilizing features deep learning methods, we present method detecting model uses pulmonary computed tomography images differentiate pneumonia from healthy cases. study, 256 cases (128 COVID-19, 128 normal) are used detect early. Real 51 external also taken Iraqi hospitals validate proposed method. Segmentations lung infection fields retrieved during preprocessing. total accuracy obtained results 98.70%, indicating success designed model.