DETECTION OF LUNG DISEASES FROM COLORIZED CHEST X-RAY IMAGES USING DEEP LEARNING

Abstract

In recent years, there have been significant advancements in the utilization of machine learning, incorporating data mining and deep learning techniques, for the analysis of chest X-ray images. These methods play a vital role as decision support tools, aiding radiologists in expediting the diagnostic process. Chest X-ray (CXR) images have proven their value in diagnosing and monitoring various pulmonary diseases, such as COVID-19 and Pneumonia and Tuberculosis. This study aims to detect these lung diseases by applying deep learning method. To achieve this, we applied Convolutional Neural Network (CNN) and Transfer (VGG16) models in the publicly available dataset comprising 7135 CXR images. The obtained results show the effectiveness of deep learning in detecting lung diseases, as well as the importance of coloring CXR images to increase the accuracy of disease detection.