부산대학교 도서관

One of the leading causes of death worldwide due to stage diffusion and delayed detection is lung cancer. Precise stage classifications are crucial to therapy methods and patient outcomes at this key juncture in the management of lung cancer. Conventional techniques depend on radiologists’ accurate interpretation of MRI and CT images, which takes time andis prone to error. As a result, algorithms based on artificial intelligence (AI) have just surfaced as a potential tool for classifying and detecting lung cancer stages to increase diagnostic consistency and accuracy. The research is divided into multiple phases, commencing with the preprocessing and data retrieval (§3). The dataset utilizes lung cancer stage labels from a few popular medical imaging databases to create a sort of overall MRI/CT information collection. Preprocessing: To normalize the images for feature extraction, this procedure entails noise removal, contrast improvement, and realization. After that, the photos are used to train Convolutional Neural Networks (CNN), whose layers alter the input image in several ways to teach them how to identify distinguishing features in a hierarchical fashion. The architecture can explain important characteristics unique to lung tumors, such as thumbnails, irregular boundaries, and cavitations, which set them apart from other types of cancer, by utilizing lung-specific layers.