Improving 3D Imaging with Pre-Trained Perpendicular 2D Diffusion Models
This addresses the challenge of exploiting 3D distribution priors in diffusion-based methods for medical imaging applications, offering a solution to improve 3D reconstruction quality.
The paper tackles the problem of 3D inverse problems in medical imaging by proposing a novel approach using two perpendicular pre-trained 2D diffusion models to model 3D data distributions, effectively addressing the curse of dimensionality. The method demonstrates effectiveness in 3D medical image reconstruction tasks such as MRI Z-axis super-resolution, compressed sensing MRI, and sparse-view CT, generating high-quality voxel volumes.
Diffusion models have become a popular approach for image generation and reconstruction due to their numerous advantages. However, most diffusion-based inverse problem-solving methods only deal with 2D images, and even recently published 3D methods do not fully exploit the 3D distribution prior. To address this, we propose a novel approach using two perpendicular pre-trained 2D diffusion models to solve the 3D inverse problem. By modeling the 3D data distribution as a product of 2D distributions sliced in different directions, our method effectively addresses the curse of dimensionality. Our experimental results demonstrate that our method is highly effective for 3D medical image reconstruction tasks, including MRI Z-axis super-resolution, compressed sensing MRI, and sparse-view CT. Our method can generate high-quality voxel volumes suitable for medical applications.