Prediction of Lung Cancer Using FSSO Optimization and Deep Learning based CNN Algorithm

Abstract

Lung cancer is the uncontrollable growth of cells within the lungs. Commonly cancer occurs in both male and female. This causes a significant breathing problem in both inhaling and exhales a part of the chest. Earlier identification of cancer is the best way for reducing mortality rate. And also identification of lung cancer is the challenging task for every researchers and doctors. In this study, we proposed the deep learning based convolutional neural network (CNN) for classification and prediction of lung cancer in earlier stage. By improving the classification accuracy, we using different significant methods as averaging histogram equalization (AVHEQ) for pre-processing to enhance and remove the noise from the dataset image , social spider optimization algorithm (FSSO) for segment the cancer affected region and extraction technique for extracted the affected region from entire image. Finally the deep learning classifier model to predict the lung cancer in earlier stage. In this model, LIDC-IDRI dataset is used to analyse the results. And also we compare the different machine learning classifier model with deep learning method. The performance of the proposed model is calculated by using different parametric measures. Finally the deep learning model achieve the better classification accuracy than other compared machine learning classifier.