Improving 3D Imaging with Pre-Trained Perpendicular 2D Diffusion Models

TL;DR

This paper introduces a novel 3D imaging method that leverages two perpendicular pre-trained 2D diffusion models to improve 3D medical image reconstruction, effectively addressing high-dimensional challenges.

Contribution

The paper proposes using two perpendicular 2D diffusion models to model 3D data as a product of 2D distributions, enhancing 3D inverse problem-solving.

Findings

Effective for MRI Z-axis super-resolution

Improves compressed sensing MRI reconstruction

Generates high-quality 3D voxel volumes

Abstract

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.