Lung Disease Diagnosis Using Various Deep Learning Algorithms from the chest X-ray images

Over the past last years, computer vision and deep learning (DL) have completely changed the world. So-called deep neural networks, which capitalize on advancements in processing power and data availability, have dominated the area of DL. The CNN method for the identification of lung diseases, including COVID-19, tuberculosis, viral pneumonia, and bacterial pneumonia, is used in this study. This study uses the EfficientNetB1, ResNet50, and VGG19 architectures in a big data setting. The Apache Spark environment is used to develop the system. The performance of EfficientNetB1, RestNet50, and VGG19 architectures to detect diseases in chest radiographs dataset containing 18383 images Chest X-ray (CX) to forecast the model's performance. Two datasets are created by splitting down the dataset. When the model's performance was evaluated and contrasted with other models, it produced results with a high f1-score, accuracy, and precision. We have attained 99% accuracy in our work.