Diagnosis of Lung Diseases From Radiography Images Using Deep Learning Methods

Nowadays, huge amounts of image data are being created for use in various fields. The use of these data in various fields such as medicine is of great importance. Thanks to the use of image analysis in the field of medicine for many years, there have been considerable improvements in the detection and treatment of diseases. In this study, Convolutional Neural Networks (CNN) which is one of the deep learning techniques has been utilized to extract distinctive features from radiography images. The classification of lung diseases has been implemented by using deep learning and machine learning models from these features. In this context, steps such as data type convertion, normalization and digitizing class information are used as pre-processing. While CNN's multilayer perceptron algorithm has been applied within the scope of deep learning, SVM and XGBoost algorithms have been implemented within the scope of machine learning. When the success of all methods has been analyzed, in multi classification the highest accuracy rate was achieved by %92,2 with the use of CNN and SVM together.