Abstract:Fiber orientation distributions (FODs) is a popular model to represent the diffusion MRI (dMRI) data. However, imaging artifacts such as susceptibility-induced distortion in dMRI can cause signal loss and lead to the corrupted reconstruction of FODs, which prohibits successful fiber tracking and connectivity analysis in affected brain regions such as the brain stem. Generative models, such as the diffusion models, have been successfully applied in various image restoration tasks. However, their application on FOD images poses unique challenges since FODs are 4-dimensional data represented by spherical harmonics (SPHARM) with the 4-th dimension exhibiting order-related dependency. In this paper, we propose a novel diffusion model for FOD restoration that can recover the signal loss caused by distortion artifacts. We use volume-order encoding to enhance the ability of the diffusion model to generate individual FOD volumes at all SPHARM orders. Moreover, we add cross-attention features extracted across all SPHARM orders in generating every individual FOD volume to capture the order-related dependency across FOD volumes. We also condition the diffusion model with low-distortion FODs surrounding high-distortion areas to maintain the geometric coherence of the generated FODs. We trained and tested our model using data from the UK Biobank (n = 1315). On a test set with ground truth (n = 43), we demonstrate the high accuracy of the generated FODs in terms of root mean square errors of FOD volumes and angular errors of FOD peaks. We also apply our method to a test set with large distortion in the brain stem area (n = 1172) and demonstrate the efficacy of our method in restoring the FOD integrity and, hence, greatly improving tractography performance in affected brain regions.

What problem does this paper attempt to address?

This paper aims to address the problem of fiber orientation distribution (FOD) recovery in diffusion magnetic resonance imaging (dMRI) under high distortion conditions. Specifically, the article points out that in dMRI, distortions caused by magnetic field inhomogeneities lead to signal loss, which affects the quality of FOD reconstruction and hinders fiber tracking and connectivity analysis in affected brain regions (such as the brainstem). Although various distortion correction methods exist, their effectiveness is limited in areas with strong distortions. To overcome this challenge, the authors propose a new method based on a diffusion model—the FOD-Diffusion model. This model can recover signal loss from severely distorted dMRI data and generate high-quality FOD images. By encoding volume sequences and extracting cross-attention features across all spherical harmonics (SPHARM) orders, the model enhances the ability to generate individual FOD volumes and maintains the geometric consistency of the generated FOD. Additionally, the model uses FOD from surrounding low-distortion areas as conditional input to maintain the geometric coherence of the generated FOD. Experimental results show that this method achieves significant success on the UK Biobank dataset, not only demonstrating high accuracy in the test set but also effectively improving fiber tracking performance in brain regions affected by distortion.

Diffusion Model-based FOD Restoration from High Distortion in dMRI