De-Noising-of-Low-Dose-CT-images-using-Generative-Models

De-Noising of Low-Dose CT images using Generative Models: Exposing radiation to patients can cause medical hazards. But the low radiations will result in poor quality of the CT Scan images due to lesser penetration into the tissues. Thus, we approach this problem by denoising the low-quality CT scans generated by low-dose radiation using various generative models such as the GAN-based Wasserstein-GAN and Pix2Pix GAN, and Auto-encoder-based RED-CNN. We used the real clinical dataset of normal dose CT Scan images which was released for “the 2016 NIH-AAPM-Mayo Clinic Low Dose CT Grand Challenge” by the Mayo Clinic. The low-dose CT scan images are simulated by introducing the combination of Gaussian electronic noise and Poisson quantum noise.