Diagnosis of COVID-19 from chest X-ray images using wavelets-based depthwise convolution network

Coronavirus disease 2019 also known as COVID-19 has become a pandemic. The is caused by beta coronavirus called Severe Acute Respiratory Syndrome 2 (SARS-CoV-2). severity of the can be understood massive number deaths and affected patients globally. If diagnosis fast-paced, controlled in better manner. Laboratory tests are available for diagnosis, but they bounded testing kits time. use radiological examinations that comprise Computed Tomography (CT) used disease. Specifically, chest X-Ray images analysed to identify presence patient. In this paper, an automated method from proposed. presents improved depthwise convolution neural network analysing images. Wavelet decomposition applied integrate multiresolution analysis network. frequency sub-bands obtained input fed identifying designed predict class image normal, viral pneumonia, COVID-19. predicted output model combined with Grad-CAM visualization diagnosis. A comparative study existing methods performed. metrics like accuracy, sensitivity, F1-measure calculated performance evaluation. proposed than methodologies thus effective