Deep learning based computed tomography image classification of COVID-19 patients

Abstract

Accurate identification of the newest 2019 coronavirus (COVID-19) disease is required for effective illness behavior and management. To categorize and analyses COVID-19 in the prevalent region, for (COVID-19) detection using medical images, computed tomography (CT) imaging is informative, reliable, and quick. Chest CT images are readily available in nearly all hospitals, making it possible to use them to classify COVID-19 patients early on. When COVID-19 infection spreads quickly, a significant amount of time are required for the chest CT-based COVID-19 categorization. Since medical practitioners have scarce time, a computerized analysis of CT scans is required. In this paper, we construct a classification framework involving the extraction of CT image characteristics and categorization. The framework is separated into training and testing modules, with training the classifier aiding in the development of a model that efficiently classifies CT images as input. Deep convolutional neural network including VGG16, ResNet50, DensNet121, and inspectionResNetV2, variations were utilized as classifiers in the present investigation (DCNN) different machine learning and Statistical modeling techniques to identify COVID-19 infections probability and discover lesion for training.