Denoising Diffusion Models for 3D Healthy Brain Tissue Inpainting

TL;DR

This paper develops and evaluates advanced denoising diffusion models for inpainting healthy brain tissue in 3D MRI scans, enabling better analysis of pathological scans by restoring healthy tissue regions.

Contribution

It introduces and compares multiple 3D diffusion-based inpainting models, including pseudo-3D and latent space approaches, with the pseudo-3D model showing superior performance.

Findings

Pseudo-3D model achieves best structural similarity and image quality metrics.

Fine-tuned model outperforms FSL lesion-filling in segmentation tasks.

Diffusion models effectively synthesize healthy tissue in pathological brain images.

Abstract

Monitoring diseases that affect the brain's structural integrity requires automated analysis of magnetic resonance (MR) images, e.g., for the evaluation of volumetric changes. However, many of the evaluation tools are optimized for analyzing healthy tissue. To enable the evaluation of scans containing pathological tissue, it is therefore required to restore healthy tissue in the pathological areas. In this work, we explore and extend denoising diffusion models for consistent inpainting of healthy 3D brain tissue. We modify state-of-the-art 2D, pseudo-3D, and 3D methods working in the image space, as well as 3D latent and 3D wavelet diffusion models, and train them to synthesize healthy brain tissue. Our evaluation shows that the pseudo-3D model performs best regarding the structural-similarity index, peak signal-to-noise ratio, and mean squared error. To emphasize the clinical relevance, we fine-tune this model on data containing synthetic MS lesions and evaluate it on a downstream brain tissue segmentation task, whereby it outperforms the established FMRIB Software Library (FSL) lesion-filling method.