Effective lung cancer detection using deep learning network

Abstract

The use of a computer-assisted diagnosis system was crucial to the results of the clinical study conducted to determine the nature of the human illness. When compared to other disorders, lung cancer requires extra caution during the examination process. This is because the mortality rate from lung cancer is higher because it affects both men and women. Poor image resolution has hampered previous lung cancer detection technologies, preventing them from achieving the requisite degree of dependability. Therefore, in this study, we provide a unique approach to lung cancer prognosis that makes use of improved machine learning and processing of images. Images of lung disease from CT scan databases created using quasi cells are used for diagnosis. Multilayer illumination was used to analyse the generated images, which improved the precision of the lungs' depiction by probing each and every one of their pixels while simultaneously decreasing the amount of background noise. Lung CT images are pre-processed to remove noise, and then a more advanced deep learning network is used to isolate the affected region. The territory is partitioned into subnetworks according to the number of existing networks, from which different features are subsequently extracted. Next, an ensemble classifier should be used to correctly diagnose lung diseases. Using MATLAB simulation, the authors examine how the provided technique improves the rate at which lung cancer could be diagnosed.