Resoluteneuronet: Deep Learning-based Segmentation and Classification Covid-19 Using Chest X-ray Images

Abstract The worldwide healthcare systems are facing substantial problems because of impending COVID-19 pandemic epidemic, necessitating rapid, accurate diagnosis for effective management and control. Chest X-ray (CXR) imaging is a valuable diagnostic tool for identifying COVID-19-related lung abnormalities. However, manual interpretation of CXR images was time-consuming and prone to inter-observer variability. Thisstudy proposes a DL approach to address the limitation of automated COVID-19 segmentation and classification utilizing CXR images. The approach segments and classifies COVID-19. The distinction between COVID-19 and healthy patients in this study was made using CXR images. Using the histogram equalization technique, the gathered CXR images are preprocessed. We can obtain images of the lungs by utilizing the “conditional generative adversarial network” (C-GAN) to segment the raw CXR images. Next, significant points were extracted from the segmented lung pictures using the Wavelet Transform(WT) to remove discriminatory features. In this study, we developed a classification method called ResoluteNeuroNet (Resolute-NN), used in the final stage to classify COVID-19 and standard lung images. The accuracy of our proposed model's performance was compared to that of the currently used COVID-19 detection methods. The experimental findings for our suggested practice demonstrate Resolute-NN superiority over the presently used state-of-the-art approaches.