Deep Learning and Patch Processing Based Lung Cancer Detection on CT Images

Abstract

Lung cancer is the most typical cancer in both men and women. Lung cancer symptoms typically appear in the human body when it is in its final stage, but with the aid of sophisticated technology and computer-aided systems, it is possible to predict at initial phase. Presently, a variety of conventional and machine learning techniques are available but they do not provide accurate detection and also takes a long time to process lung cancer detection. As a result, a novel method for detecting lung cancer in CT images was proposed which employs a patch processing and deep learning techniques for accurate detection and less computation time. The image quality of CT scans are improved by patch processing and deep learning technique classifies the detected tumor as malignant or benign. Finally, the statistical parameters such as Accuracy, MSE, Sensitivity, PSNR, Sensitivity, and Specificity were computed and compared with the existing system.