Detection and Staging of Lung Cancer from CT scan Images by Deep Learning

Lung cancer is jeopardized disease which easily demolish human body with great impact. Lung Cancer is most prominent disease nowadays. To diagnosis this minacious disease CT scan is one of the methods which is more often used. Manual detection of anomalies from medical images It takes a lot of time and is prone to aberration. Several computer-aided diagnosis methods have been developed for the accurate diagnosis of cancers from images. With the advent of deep learning approaches in image processing, researchers have moved to developing new and efficient models for cancer diagnosis and classification. In this work, a convolution neural network-based approach has been presented for the prediction and categorization of cancers from lung CT scans. After image preprocessing and resizing, the class imbalance is taken care of by augmentation. Batch normalization, max pooling, and flattening are some of the steps used in the model. Finally, fully connected layer functioned to classify the images into four classes, namely benign, primary lung cancer, malignant, and suspicious malignant. The results obtained are comparable to the ones achieved by other researchers so far.