Real-Time Lung Cancer Detection Using GAN-Enhanced LDCT Imaging and CNN-Based Classification for Clinical Accuracy

Lung cancer stands as the leading contributor to cancer- related casualties on a global scale, holds the topmostmortality rates for both males and females. Tobacco using up is responsible for approximately equivalent to 85% of all instances of lung cancer, which makes it the primary cause of the illness. Frequently, cancer of the lungs is suspected in its advanced stages, limiting treatment options. Lung cancer is still one of the leading causes of cancer- related deaths globally, and better patient outcomes depend on early identification. Low- Dose Computed Tomography (LDCT) scans are widely utilized to test for lung cancer, but the inherent noiseand Variations in raw LDCT images pose substantial challenges for accurate diagnosis. This paper proposes a real-time lung cancer detection system that leverages the power of Convolutional Neural Networks (CNN) and Generative Adversarial Networks (GAN). The system is designed to classify lung cancer in LDCT images while addressing the issue of image quality degradation inraw scans. The GAN model is employed to generate high quality synthetic images from raw LDCT inputs, enhancing their clarity and usability for diagnostic purposes. These improved images are then fed into a CNN model, which performs the lung cancer detection. The proposed system was assessed and evaluated on a dataset of LDCT images, demonstrating superior effectiveness in classification sensitivity, specificity, and accuracy relative to traditional methods. Our results highlight the potentialof combining GAN - generated synthetic images with CNN-based classification for more accurate and reliable real-time lung cancer detection, providing a robust solution for early-stage cancer diagnosis in clinical settings.