Semantic Literature Analysis and Deep Learning Algorithm Comparison for Identifying Paru-Paru Diseases from X-rays

Abstract

The rapid development of artificial intelligence, particularly deep learning, has significantly improved the performance of medical image analysis, including chest X-ray interpretation for lung disease detection. This study presents a semantic literature analysis combined with a comparative review of deep learning algorithms used to identify pulmonary diseases from chest X-ray images. Rather than proposing a novel algorithm, this research systematically examines and synthesizes existing studies that employ convolutional neural networks (CNN), transfer learning models, and hybrid deep learning architectures for detecting lung-related conditions such as pneumonia, tuberculosis, COVID-19, and lung cancer. The semantic analysis approach is applied to identify research patterns, dominant methodologies, datasets, evaluation metrics, and performance trends across selected publications. Furthermore, a comparative analysis is conducted to evaluate the strengths and limitations of commonly used deep learning models based on accuracy, sensitivity, specificity, and computational efficiency. The findings reveal that transfer learning-based CNN models generally outperform traditional architectures, particularly when applied to limited datasets. This study provides a comprehensive overview of current research trends and offers valuable insights for researchers and practitioners in selecting appropriate deep learning approaches for lung disease identification using X-ray images.