Automated Classification of Pneumonia from Chest X-Ray Images using Deep Transfer Learning EfficientNet-B0 Model

Pneumonia is an infectious and fatal respiratory disease. A radiologist needs time and resources to diagnose or assess pneumonia from a chest x-ray image. Detecting a disease might occasionally be difficult due to noise of experimental and clinical data being acquired utilizing a variety of techniques from various research laboratories. As a result of this challenge, numerous deep learning frameworks have been proposed, with promising results as the accuracy of positive case detection increases. This study uses the CNN-based transfer learning model EfficientNet-B0, which is then tweaked and trained using a publicly available dataset, based on the previous work on automated x-ray image screening for pneumonia identification. Techniques for data preparation and augmentation are also used to address the issue of class imbalance. Overall accuracy, precision, recall, F-score, and area under the ROC curve for the model are 92.6%, 92.3%, 91.8%, 92.1% and 91.9%, respectively.