Abstract: Lung cancer is one of the most dangerous cancers. Thousands of people are infectedwith it, and if they do not detect it in the early stages of the disease, the patient's chances ofsurvival are slim to none. For the reasons stated above, and to aid in the fight against this dreadfuldisease, early detection via artificial intelligence processes is necessary. Using a convolutionalneural network technique with proposed architecture, a computer-aided Lung Cancer Detectionsystem(LCDS) system is proposed for detecting lung cancer in human lungs by using computedtomography (CT scan) images for assisting patient with the diagnosis of the lung cancer in threestages normal, beginning or malignant. The proposed model is expected to have accuracy about89.45 percent.Keywords: Dataset, CNN (Convolutional Neural Network), ReLU, Pooling, Fully-ConnectedLayer.