WDM: 3D Wavelet Diffusion Models for High-Resolution Medical Image Synthesis

arXiv:2402.19043v1 Announce Type: cross
Abstract: Due to the three-dimensional nature of CT- or MR-scans, generative modeling of medical images is a particularly challenging task. Existing approaches mostly apply patch-wise, slice-wise, or cascaded generation techniques to fit the high-dimensional data into the limited GPU memory. However, these approaches may introduce artifacts and potentially restrict the model's applicability for certain downstream tasks. This work presents WDM, a wavelet-based medical image synthesis framework that applies a diffusion model on wavelet decomposed images. The …

abstract apply arxiv cs.cv data diffusion diffusion models eess.iv generative generative modeling gpu image images medical memory modeling nature scans synthesis three-dimensional type wavelet wise