Ground-truth effects in learning-based fiber orientation distribution estimation in neonatal brains

TL;DR

This study compares ground-truth FOD estimation methods for neonatal brains, finding that single-shell three-tissue CSD provides more realistic and robust results than multi-shell methods, especially with increased gradient directions.

Contribution

It demonstrates that SS3T-CSD is more suitable than MSMT-CSD as ground truth for neonatal FOD estimation, especially under age domain shifts.

Findings

SS3T-CSD yields more realistic fiber voxel ratios.

Increasing gradient directions improves SS3T-CSD performance.

SS3T-CSD maintains robustness across different neonatal ages.

Abstract

Diffusion Magnetic Resonance Imaging (dMRI) is a non-invasive method for depicting brain microstructure in vivo. Fiber orientation distributions (FODs) are mathematical representations extensively used to map white matter fiber configurations. Recently, FOD estimation with deep neural networks has seen growing success, in particular, those of neonates estimated with fewer diffusion measurements. These methods are mostly trained on target FODs reconstructed with multi-shell multi-tissue constrained spherical deconvolution (MSMT-CSD), which might not be the ideal ground truth for developing brains. Here, we investigate this hypothesis by training a state-of-the-art model based on the U-Net architecture on both MSMT-CSD and single-shell three-tissue constrained spherical deconvolution (SS3T-CSD). Our results suggest that SS3T-CSD might be more suited for neonatal brains, given that the ratio between single and multiple fiber-estimated voxels with SS3T-CSD is more realistic compared to MSMT-CSD. Additionally, increasing the number of input gradient directions significantly improves performance with SS3T-CSD over MSMT-CSD. Finally, in an age domain-shift setting, SS3T-CSD maintains robust performance across age groups, indicating its potential for more accurate neonatal brain imaging.