EFFICIENT CONVOLUTIONAL NEURAL NETWORK FOR THE EARLY PREDICTION OF LUNG CANCER

Computer Engineering

One of the most common illnesses in the world's population and the main cause of the increasing mortality rate is lung cancer. Medical experts believe computed tomography (CT) screening for lung cancer early detection can reduce mortality in patients with lung nodules. Lung cancer killed 40,000 people annually in the United States in 1965; by 2000, that number had risen to almost 200,000. Overall, the 5-year survival rate is still around 15%. It is only recently that the incidence of lung cancer has started to reduce, primarily due to a decrease in smoking. Tobacco use is thought to be a risk factor for a number of malignancies in addition to lung cancer, such as leukemia and cancers of the mouth, throat, bladder, ureter, kidney, esophagus, stomach, pancreas, and throat. This study presents an effective CNN for the early detection of lung cancer, together with Mayfly optimization and CNN classification techniques for lung cancer prediction. Lung tissue must be taken beforehand to decrease the detection area, save computing time, and increase accuracy. A better maximal inter-class variance (OTSU) technique that incorporates morphological processes is put forward. CNN are a type of artificial neural network that are mostly used for image processing and identification since they are able to recognize patterns in pictures. The architecture of the CNN model is built utilizing the parameters derived from the Grey Level Co-occurrence Matrix (GLCM) technique for feature extraction. CNN has an accuracy of 90.7% and a specificity comparison of 91%.

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.