R-GAN-NET: A Novel Deep Convolution Neural Network with Augmentation to Detect Lounge Nodules
The early identification of lung nodules is important for predicting the development of lung cancer. Lung nodules have a high prospect of turning cancerous, and spotting malignant nodules early improves a patient's recovery prospects. Diagnostic methods currently rely on manual assessment and machine learning, particularly neural network techniques that consider nodule size and color. However, these methods may not be very influential in detecting cancer in its early stages, often taking months or even years to confirm malignancy. Detecting malignant nodules early can significantly extend a person's life. To address this challenge, An R-GAN-NET model proposed an innovative approach that utilizes a R-GAN and deep neural network models to determine whether lung nodules are malignant. For this LIDC-IDRI Kaggle data set is used to train the model. The method effectively deals with small and overlapping nodules by incorporating a Deep Convolutional Network to analyze images and carefully recognize distinct patterns in nodules. The proposed model identifies the malignant nodules without concern for the size or color of the samples. And achieved exceptional performance compared to the prescribed standards, with an accurateness of 96.1%.