One of the main causes of cancer-related fatalities globally is lung cancer, and the prognosis for patients is greatly enhanced by early identification. Recent developments in deep learning have shown great promise for automating the examination of medical images. Using the publicly accessible LIDC-IDRI dataset, this study suggests a convolutional neural network (CNN) and a transfer learning-based ResNet-50 architecture for the identification and categorization of lung nodules. The dataset consists of pre-processed annotated thoracic CT scans that were used to train and assess the suggested models. With an accuracy of 93.2% and an AUC of 0.96, our experimental results demonstrate that the ResNet-50 model performed better than the baseline CNN, demonstrating its suitability for clinical use. The study also talks about the problems, restrictions as well as the prospects to enhance the diagnosis of lung cancer through artificial intelligence in the future.