Authors: Peng Wang; Zijuan Zhao
Addresses: ShangXi Eye Hospital, Taiyuan 030002, China ' College of Information and Computer, Taiyuan University of Technology, Taiyuan 030024, China
Abstract: Classification of benign and malignant pulmonary nodules is a critical task for developing a Computer-Aided Diagnosis (CAD) system of lung cancer. However, the intelligent diagnosis technology is often limited by equipment and space. Therefore, a computer-aided diagnosis model is proposed running in the Mobile Edge Computing (MEC) environment. The novel lung nodule classification framework improved the Deep Convolutional Generative Adversarial Nets (DCGAN). Firstly, CT images after pre-processing are input into GAN to generate new images with similar features. Then, in the training stage, the derivative model of GAN is introduced into the classification of pulmonary nodules. The optimised function means the improved DCGAN has better anti-noise ability and achieves more accurate classification performance. The experimental results showed an accuracy of 88.88%. Compared with the existing methods, the proposed method is superior to other ones in terms of accuracy, sensitivity, specificity, and area under the ROC curve.
Keywords: pulmonary nodules; generative adversarial nets; feature extraction; classification; mobile edge computing.
International Journal of Wireless and Mobile Computing, 2020 Vol.18 No.1, pp.80 - 89
Received: 31 Aug 2019
Accepted: 23 Nov 2019
Published online: 28 Jan 2020 *