Deep Learning Model-based Approach for Lung Cancer Detection

Cancer disease detection is a critical task in healthcare systems when different technologies or tools are used to detect the exact disease in a patient. Although various cancer diseases are considered to be analyzed in the healthcare system, lung cancer disease is one of them. A hazardous cancer disease must be detected earlier to save patient life. Traditional methods for lung cancer detection are very tedious. Thus, this paper proposed a deep learning (DL) model with various techniques to design individual models and experiments accordingly. Different techniques such as ResNet 152, DenseNet 169, Xception, and Autoencoders are considered for testing lung cancer disease data set through computed tomography (CT) images. The proposed models have demonstrated impressive performance across various image identification tasks. The model trains various algorithms using a benchmark lung cancer dataset. The experimental comparative analysis is considered for better performance between the proposed and existing models, which helped to determine the appropriate method with the correct model. The DenseNet 169 model has better performance in both training accuracy (94.60%) and testing accuracy (92.00%).