Diffusion Models for MRI generation

Obtaining MRI data in medicine is challenging due to privacy concerns, such as patient confidentiality, and logistical issues like limited access to specialized imaging equipment or variability in data acquisition protocols. On the other hand, deep learning models thrive on large datasets, and their performance often suffers in data-scarce environments. To address this, this project aims to enhance MRI-based deep learning models by generating synthetic MRI images using diffusion models, which allow us to simulate diverse, high-quality medical images while preserving patient privacy.