Abstract :
Coronavirus has affected daily lives of people all around the globe. Lungs being the respiratory organ are the most affected by such a virus. Alternative techniques for diagnosing the coronavirus involving X-rays and CT scans of the chest have been proposed. The severity of the disease, on the other hand, is a crucial component in the patient's treatment. As a consequence, an automated approach to ascertain the severity of the coronavirus on the lungs is designed to decrease the impacts of the coronavirus on the lungs and practice the right treatment. In this manuscript, we proposed a deep learning-based model for identifying the severity level of coronavirus on the lungs which is further categorized in high, moderate, and low. We employed AlexNet for the disease detection and Faster RCNN for the severity level prediction based on the affected area of the lungs. The evaluation is assessed using X-rays and CT scans of the lungs. Total 1400 images have been employed for the training and performance evaluation of the proposed system. The metrics that we considered for the performance evaluation are accuracy, precision, recall, error rate, and time. The results showed that our proposed model attained about 98.4% accuracy and 98.15% precision.
Keyword :
Classification, Detection, Deep Learning, Severity Detection, Disease Identification.