Cancer which begin to grow uncontrollably in the lungs is known as lung cancer. Lung cancer is second deadliest disease in the world. In this paper, we have implemented convolutional neural network model for detection of cancerous nodules from CT (computed tomography) scans of different infected patients. for experimentation dataset is collected from cancerimagingarchive website and performed preprocessing steps to make better fit for our model. further we split dataset into 70% training and 30% testing. we compare performance of our CNN(convolutional neural network) model with pretrained models like VGG16.at the end we have concluded with challenges,research gaps and future scope of our system. © 2022 IEEE.