Lung Nodule Detection through CT Scan Images and DNN Models

This paper focuses on detecting the nodules present in the lung CT scan images. The primary goal is to detect the lung cancer at the early stage. The step for detecting will follow image pre-processing, segmentation, data preprocessing, data augmentation, model construction, network layers. For proper image processing results, data pre-processing, data augmentation will play a major role. Here construct 2D SqueezeNet and 2D MobileNet Models for detecting the lung cancer in the patients. For classifying the lung nodules, the basic approach is Radiomics which extracts the features of the image. Another approach is using CNN models for extracting the 2D or 3D images. Here by applying deep learning model on the Lung Nodule (LUNA16) dataset consisting of the CT scan 3D Images of the patients. Constructing the architecture of the deep learning networks for efficient working of the model. The paper involves the use of DNN as it reduces the use of memory and it is a light-weight model.