Published Paper Details:

This study aims to detect lung cancer at its early stages and evaluate the accuracy of various machine learning models used in this process. Through an extensive review of existing literature, it was observed that some classification algorithms yield lower accuracy, while others perform better but still fall short of achieving near-perfect results. One of the major challenges is the high implementation cost and low accuracy, often due to the improper handling of DICOM images. In medical imaging, several types of scans are available, but CT (Computed Tomography) images are preferred due to their reduced noise levels. Deep learning has emerged as the most effective approach for tasks like lung nodule detection, classification, feature extraction, and predicting the stage of lung cancer. The proposed system initially applies image processing methods to isolate the lung region. Segmentation is carried out using the K-Means clustering algorithm. After segmentation, features are extracted and classified using multiple machine learning techniques. The efficiency of these methods is assessed based on accuracy, sensitivity, specificity, and classification time.