We are thrilled to announce that Hongxu Jiang, a student researcher in the MIRTHAI Lab, has published his latest work, “Fast-DDPM: Fast Denoising Diffusion Probabilistic Models for Medical Image-to-Image Generation,” in the prestigious IEEE Journal of Biomedical and Health Informatics (J-BHI).
IEEE J-BHI is one of the leading journals at the intersection of engineering, computer science, and healthcare, boasting a strong impact factor of 6.7. The journal is internationally recognized for publishing high-quality research that advances the development of innovative technologies for improving human health.
In his paper, Hongxu introduces Fast-DDPM, a novel diffusion model that significantly accelerates both training and sampling processes—cutting training time by up to 5× and sampling time by up to 100×—while achieving state-of-the-art performance across a variety of medical imaging tasks, including multi-image super-resolution, image denoising, and image-to-image translation. By optimizing time-step usage and designing task-specific noise schedulers, Fast-DDPM addresses key computational challenges that have historically limited the clinical application of diffusion models.
This exciting work represents an important step forward in making AI-driven solutions more practical and efficient for real-world healthcare applications.
The full article will be available soon through IEEE Xplore. Manuscript currently available on arXiv.
Congratulations to Hongxu on this outstanding achievement!