<p>Increased survival depends on prompt identification of lung cancer because it remains among the primary cancer causes of death globally. Low-dose computed tomography (LDCT) provides patients with a key advancement in early-stage diagnosis methods. However, raw imaging data can contain excessive noise, poor contrast, and other artifacts, thus it is essential to employ efficient pre-processing methods to promote better diagnostic precision. In this paper, 50 studies on pre-processing methods for lung cancer detection are reviewed. The main methods include data augmentation, lung region segmentation, edge detection, contrast stretching, image normalization, and denoising. The methods enable better tumor detection while improving image quality as well as providing better features needed for machine learning and deep learning algorithm processing. Despite these advancements, problems with false positives, overfitting from supplemented data sets, and processing costs still exist. Research should be done to examine the adaptive enhancing potential of hybrid pre-processing frameworks that incorporate artificial intelligence. Automated lung cancer detection has advanced thanks in large part to pre-processing images to enhance their quality, which is necessary for a quicker and more precise diagnosis.</p>