Metal artifact reduction in computed tomography images based on developed generative adversarial neural network

•This study shows that GAN) Networks will give the best images by considering image quality metrics.•Simulated images of head and neck were our validation of GAN success in metal artifact reduction.•Quality image metrics were compared between noisy, healthy and denoised simulated images to show the image improvement.•Patients images with one dental implant have more improvement in oral cavity area which is very important in treatment planning.