Detecting Current Lung Diseases and COVID-۱۹ from CT Images Using An Expansive Deep Learning Method

The COVID-۱۹ not only has the most rapid worldwide spread rate among other known viruses but also significantly impacts everybody’s life these days. It is challenging to detect Coronavirus disease because the vital signs are the same as those of other lung diseases, especially the Coronavirus disease family. Computer-aided diagnosis (CAD) increases diagnosis efficiency, helping doctors provide a quick and confident diagnosis and treatment of COVID-۱۹. Using CT images is one of the effective ways to detect and quantify the disease. In this study, we detect three different pandemic lung diseases from each other and propose the novel methodology to use the Convolutional Neural Network (CNN), a representational model faster than traditional algorithms to recognize the type of lung disease by CT images. The dataset has been chosen from ۳ types of pandemic lung disease comprise COVID-۱۹, Influenza (N۱H۱), and Cancer. The utilized method has high accuracy at about ۹۹.۸% with a short running time of ۲.۳۸s, which is the best one during several tests. This study may assist clinicians in reducing the missed diagnosis during COVID-۱۹ detection.

کلیدواژه ها:

Corona Virus Disease (COVID-۱۹) ، Lung Diseases ، Computerized Tomography (CT) ، Deep Learning ، Convolutional Neural Network (CNN)