GAN‐LSTM‐3D: An efficient method for lung tumour 3D reconstruction enhanced by attention‐based LSTM

Three-dimensional (3D) image reconstruction of tumours can visualise their structures with precision and high resolution. In this article, GAN-LSTM-3D method is proposed for 3D lung cancer from 2D CT images. Our consists three phases: segmentation, tumour reconstruction. Lung segmentation done using snake optimisation followed by Gustafson-Kessel (GK) clustering method. The outputs GK (2D images) are fed to a pre-trained Visual Geometry Group (VGG) feature extraction. VGG used as input an attention-based LSTM which performs unpacking. output units given generator network Generative Adversarial Networks (GAN) model carry out (normal/cancerous) images quality. During training, the discriminator GAN judge outputs. authors best knowledge were first use primary contribution article. Moreover, existing studies mostly focused on brain not suitable Focusing second Evaluation LUNA data collection against methods like MC, MC + fairing etc. reveals superiority our in terms Hamming Euclidean distance metrics. Additionally, computational complexity lower compared evaluated methods.