Denoising Diffusion Models for 3D Healthy Brain Tissue Inpainting

Monitoring diseases that affect the brain's structural integrity requires automated analysis of magnetic resonance 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 probabilistic models (DDPMs) for consistent inpainting of healthy 3D brain tissue. We modify state-of-the-art 2D, pseudo-3D, and 3D DDPMs working in the image space, as well as 3D latent and 3D wavelet DDPMs, 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 synthetic multiple sclerosis lesions and evaluate it on a downstream brain tissue segmentation task, where it outperforms the established FMRIB Software Library (FSL) lesion-filling method.