Search papers, labs, and topics across Lattice.
This paper introduces SAFE-Diff, a diffusion model for synthesizing contrast-enhanced breast MRI images that incorporates scale-aware attention to capture multi-scale lesion features and feature-dispersive diffusion to handle heterogeneous enhancement patterns. The model also estimates uncertainty to improve the reliability of the synthesized images. Experiments demonstrate that SAFE-Diff generates higher fidelity images with more realistic lesion characteristics compared to existing methods, as evaluated by quantitative metrics and qualitative assessment by radiologists.
Synthesized breast MRIs can now better mimic real-world lesion complexity thanks to a diffusion model that explicitly handles multi-scale features and heterogeneous enhancement.
Synthesizing high fidelity contrast enhanced MRI is clinically valuable for safer and more efficient breast cancer screening, yet remains challenging due to complex lesion textures and heterogeneous enhancement patterns.
