DETECTION AND CLASSIFICATION OF LUNG DISEASES USING DEEP LEARNING APPROACH

ABSTRACT:

Lung diseases pose a significant global health challenge, often necessitating early detection for effective intervention. Medical imaging, particularly chest radiographs, plays a major role in the diagnosis and disease monitoring. This research explores the use of deep learning models, specifically VGG16, VGG19, MobileNetV2 and InceptionV3 in the computerized detection and classification of lung diseases from lung X-ray images. In this research, our study focused on finding the right transfer-learning algorithm for the classification of lung diseases, specifically, COVID-19, Pneumonia, Tuberculosis and Lung cancer, using lung X-ray images. The study made a comparison of these algorithms using their accuracy, F1 score, and Recall for the selected lung images of normal and the diseases. After running ten (10) experiments, VGG16 presents the best algorithm for medical image Classification with an accuracy of 92%. This was achieved through the use of the dataset from Kaggle’s public repository, which was pre-processed, and stored in the file manager/internal storage for modeling. The final outputs of the models were deployed to a Web App using Flask and Mysql Lite Database.

Keywords: Lung Disease, Classification, Detection, Machine- Learning, Deep-Learning