Göğüs X-Ray görüntülerinin AlexNet tabanlı sınıflandırılması

Abstract

— COVID-19 pandemic first broke out in December 2019 and has been affecting the world ever since. The number of COVID-19 patients is increasing rapidly in the world day by day, and it is known that the diagnosis of this disease is important for disease treatment. Chest X-ray images that are clinical adjuncts are widely used in the diagnosis of COVID-19 disease. In the study, machine learning-based models are developed using these images to reduce the workload of expert. In the data set used in the study, there are images obtained from a total of 137 COVID-19, 90 normal, and 90 pneumonia subjects. Here, 1000 image features are extracted for each image using AlexNet deep learning architecture. Afterward, the classifiers used in the study are trained using these image features. From the results, Accuracy (%), Sensitivity (%), Specificity (%), Precision (%), F1 score (%), and Matthews Correlation Coefficient (Matthews Correlation Coefficient, MCC) values of Cubic SVM that is the most successful classifier are equal to 95.27, 94.95, 97.76, 94.65, 94.79, and 0.9250, respectively.