Diagnosis of Lung Pneumonia from X-ray Film by Deep Learning

The application of engineering technology in healthcare has significantly improved diagnostic efficiency, reducing time lost in medical workflows. Despite advancements in medical imaging, the accurate and efficient classification of lung X-ray images remains a challenge due to limited datasets and complex feature extraction requirements. To address these issues. This research presents a Hybrid Deep Learning model, combining convolutional neural networks (CNNs) with generative adversarial networks (GANs) for data augmentation, attention mechanisms for feature enhancement, and employs a robust classification strategy for classification. This novel approach addresses inefficiencies in pneumonia detection and optimizes accuracy and reliability. Using a dataset of 80,000 lateral X-ray images, the Hybrid CNN model was evaluated against ResNet-50 and VGG-16 benchmarks. Key metrics, including precision, recall, F1-score, and ROC curves, demonstrated the model’s superior performance, achieving 95% accuracy with RMSProp optimization. The findings underscore the Hybrid CNN's potential to enhance radiological workflows and advance automated medical imaging, contributing to improved patient care in healthcare facilities.