A Comparative Study of Deep Learning Models for MRI to CT Image Synthesis

This study focuses on leveraging deep learning techniques to synthesize Computed Tomography (CT) images from Magnetic Resonance Imaging (MRI) data, providing an innovative solution to reduce patient exposure to ionizing radiation while maintaining diagnostic accuracy. By implementing advanced models such as Variational Autoencoder (VAE), UNET, and Adversarial Autoencoder GAN (AAE GAN), this work demonstrates the potential of deep learning in medical image translation. The synthesized CT images effectively capture critical anatomical details, offering a reliable alternative to traditional imaging methods. This work highlights the transformative role of artificial intelligence in medical imaging and its potential to enhance clinical workflows by enabling efficient and non-invasive diagnostic solutions.