ADVANCES IN INFORMATION AND KNOWLEDGE-ENGINEERING: FROM DATA TO INTELLIGENT SYSTEMS

The early and accurate diagnosis of lung diseases, in- cluding lung cancer and tuberculosis, is critical for effective treat- ment and improved patient outcomes. However, the diagnostic process is often hindered by limitations in conventional imaging techniques, such as low-dose computed tomography (LDCT) and chest X-rays, which can result in images with low contrast, noise, and artifacts. The emergence of machine learning (ML) and deep learning (DL) algorithms has revolutionized the field, offering promising solutions to enhance lung image quality and diagnostic accuracy. This comprehensive review explores the advancements in ML/DL methodologies for improving lung image quality. It delves into challenges associated with conventional imaging techniques, data scarcity, and annotation limitations, alongside discussing noise reduction, contrast enhancement, and segmen- tation techniques. The paper further evaluates ML/DL-based algorithms, including convolutional neural networks (CNNs), generative adversarial networks (GANs), and hybrid approaches, highlighting their impact on image enhancement and disease detection. By analyzing performance metrics and comparative studies, the review underscores the potential of these technologies to transform clinical workflows. The findings emphasize the need for robust datasets, enhanced generalize ability, and explain ability to facilitate widespread clinical adoption and improve patient outcomes.