Lung cancer is an important healthcare concern. It affects gender in different ways and men are at the greatest risk of dying from this cancer, while women are at the second highest risk. In medical science, the process of detection of lung cancer is done using low-dose CT scan (LDCT) images. For this purpose, a patient lies on a thin and flat table sliding back and forth inside a hole in the middle of the Computed Tomography (CT) scanner, which is a large, doughnut-shaped healthcare device. Doing the same using automated systems using artificial intelligence with high accuracy without human involvement is a challenging task. The main objective of this research is to design a Convolutional Neural Network (CNN) model to classify squamous cell carcinoma and adenocarcinoma with high accuracy. Around 85% of lung cancer cases belong to non-small cell lung cancer. Early detection and treatment are important to a patient’s recovery. Diagnosing the various kinds of cancers of the lungs is usually a troublesome process that requires time and error. In addition to identifying lung cancer subtypes more accurately and in less time, convolutional neural networks may help in determining patients’ right treatment procedures and their survival rates. Despite its complexity, even for experienced pathologists, this area of research can be challenging when it comes to adenocarcinoma and squamous cell carcinoma. This chapter proposes a mathematical model and its three-layer CNN implementation for sub-classification of non-small cell lung cancer into squamous cell carcinoma and adenocarcinoma. The proposed system is validated to classify non-small cell lung cancer into squamous cell carcinoma and adenocarcinoma. The model is trained with an accuracy of 96.89% and validated with an accuracy of 93.20%.