Identification of Coronavirus Pneumonia using a Proposed Deep Learning-based Algorithm

CT images and deep learning algorithms will be used in this study to diagnose COVID-۱۹. First, a new approach to removing noise from CT images is presented using wavelet transformation and fuzzy logic. The combined global and local threshold method was then used to segment lung images. As a result, lung regions can be successfully segmented from CT images. The next step will be the extraction of features and classifications. Feature extraction is carried out by AlexNet, while classification is carried out by Support Vector Machines (SVM). COVID-۱۹, viral pneumonia, and normal data are classified with ۹۹.۸% accuracy. This method's classification performance is superior to those of previous methods.

Keywords:

Coronavirus Pneumonia (COVID-۱۹) , Lung segmentation , Support vector machine (SVM) , Convolutional neural networks , AlexNet.