Diffusion Transformer Meets Random Masks: An Advanced PET Reconstruction Framework
Journal:
arXiv
Published Date:
Mar 11, 2025
Abstract
Deep learning has significantly advanced PET image re-construction, achieving
remarkable improvements in image quality through direct training on sinogram or
image data. Traditional methods often utilize masks for inpainting tasks, but
their incorporation into PET reconstruction frameworks introduces
transformative potential. In this study, we pro-pose an advanced PET
reconstruction framework called Diffusion tRansformer mEets rAndom Masks
(DREAM). To the best of our knowledge, this is the first work to integrate mask
mechanisms into both the sinogram domain and the latent space, pioneering their
role in PET reconstruction and demonstrating their ability to enhance
reconstruction fidelity and efficiency. The framework employs a
high-dimensional stacking approach, transforming masked data from two to three
dimensions to expand the solution space and enable the model to capture richer
spatial rela-tionships. Additionally, a mask-driven latent space is de-signed
to accelerate the diffusion process by leveraging sinogram-driven and
mask-driven compact priors, which reduce computational complexity while
preserving essen-tial data characteristics. A hierarchical masking strategy is
also introduced, guiding the model from focusing on fi-ne-grained local details
in the early stages to capturing broader global patterns over time. This
progressive ap-proach ensures a balance between detailed feature preservation
and comprehensive context understanding. Experimental results demonstrate that
DREAM not only improves the overall quality of reconstructed PET images but
also preserves critical clinical details, highlighting its potential to advance
PET imaging technology. By inte-grating compact priors and hierarchical
masking, DREAM offers a promising and efficient avenue for future research and
application in PET imaging. The open-source code is available at:
https://github.com/yqx7150/DREAM.
