Lung Cancer Prediction using Feature Selection and Recurrent Residual Convolutional Neural Network (RRCNN)

Lung cancer (LC) is the most prevailing cause of cancer deaths globally. Detecting LC by using Computed Tomography (CT) images is the predominant method. Over recent years, Deep learning (DL) techniques have a profound impact on providing the best performance in various fields of research. These techniques are successfully implemented in the medical imaging field. The U-Net model is the most used and practiced algorithm for medical imaging applications. In the present study, we proposed the recurrent Residual Convolutional Neural Network (RRCNN) model on basis of UNet models, which automatically segment and classify the lung nodules. The proposed model makes use of three underlying algorithms i.e., UNet, Residual Network, as well as Recurrent Convolutional Neural Network (RCNN). For experimental analysis, the LUNA16 dataset had been utilized and the outcome of the model demonstrates that it can efficiently accomplish tasks like detection, identification, segmentation, and classification from the input given compared to identical models along with UNet and RCNN and achieved an accuracy of around 97%.