SMGDiff: step mapping generalized diffusion model for efficient noise reduction in cardiac-gated myocardial perfusion SPECT images

Key Points

• The aim is to develop the SMGDiff model for effective denoising of cardiac-gated myocardial perfusion SPECT images.
• Developed the SMGDiff-5 model for image processing.
• Compared performance with traditional deep learning methods.
• Assessed computational efficiency during implementation.
• SMGDiff-5 demonstrated superior denoising performance.
• Significantly lower computational demand compared to conventional methods.