Self‐Supervised Deep Learning Framework for Rician Distribution Based Denoising and Modeling of Multi‐b Prostate Diffusion MRI

Key Points

• This research aims to enhance the quality of diffusion MRI images using a self-supervised deep learning framework.
• Utilized a CNN-based approach for image denoising and modeling.
• Focused on Rician distribution for analysis of multi-b prostate diffusion MRI.
• Emphasized the need for larger training datasets for better outcomes.
• Achieved higher quality diffusion-weighted images compared to traditional methods.
• Significantly reduced computation times in image processing.
• Indicated that further refinements are necessary for robustness and generalizability.