Generative Adversarial Networks (GANs) have emerged as valuable tools in deep learning for recognizing patterns in images. These patterns are used particularly for enhancing disease diagnosis. The medical imaging field has witnessed an increase in the application of GANs. However, this surge in usage has brought forth a new challenge in accurately identifying fake or manipulated medical images. Among various medical imaging modalities, lung Computed Tomography (CT) scans have become a focal point due to the potential for conditional GANs to generate deceptive images. This research study aims to address the critical issue of identifying manipulated lung CT scans. If left undetected, these manipulated images could have severe consequences for patients, including misdiagnosis, harm, or even loss of life. This new approach has been developed using a deep 3D VGG network to detect fake lung CT scans. Employing ablation study that alters the input set of layers of the neural network helps to achieve significant improvements and achieves an accuracy of 93.98% and these results indicate a notable 2% increase in performance compared to previous methods. Therefore, the findings also suggest that the model can be effectively used in detecting fake cancer nodules.