Paper

Domain shifts in infant brain diffusion MRI: Quantification and mitigation for fODF estimation

Bach Cuadra, Meritxell, Thiran, Jean-Philippe, Kebiri, Hamza, Chen, Yufei, Karimi, Davood, Lin, Rizhong, Gholipour, Ali

2025

Overview

The accurate estimation of fiber orientation distributions (fODFs) in diffusion magnetic resonance imaging (MRI) is crucial for understanding early brain development and its potential disruptions. While deep learning (DL) models have shown promise in fODF estimation from neonatal diffusion MRI (dMRI) data, the out-of-domain (OOD) performance of these models remains largely unexplored, especially under diverse domain shift scenarios. This study evaluated the robustness of three state-of-the-art DL architectures—MLP-based, Transformer-based, and U-Net/CNN-based—on fODF predictions derived from diffusion MRI data. Using 488 subjects from the developing Human Connectome Project (dHCP) and the Baby Connectome Project (BCP) datasets, we reconstructed reference fODFs from the full dMRI series using single-shell three-tissue constrained spherical deconvolution (SS3T-CSD) and multi-shell multi-tissue CSD (MSMT-CSD) to generate reference fODF reconstructions for model training, and systematically assessed the impact of age, scanner/protocol differences, and input dimensionality on model performance. Our findings reveal that U-Net consistently outperformed other models when fewer diffusion gradient directions were used, particularly with the SS3T-CSD-derived ground truth, which showed superior performance in capturing crossing fibers. However, as the number of input diffusion gradient directions increased, MLP and the transformer-based model exhibited steady gains in accuracy. Nevertheless, performance nearly plateaued from 28 to 45 input directions in all models. Age-related domain shifts were found to be less pronounced in late developmental stages (late neonates, and babies), with SS3T-CSD demonstrating greater robustness to variability compared to MSMT-CSD. To address inter-site domain shifts, we implemented two adaptation strategies: the Method of Moments (MoM) and fine-tuning. Fine-tuning consistently yielded superior results, with U-Net benefiting the most from increased target subjects. This study represents the first systematic evaluation of OOD settings in deep learning applications to fODF estimation, providing critical insights into model robustness and adaptation strategies for diverse clinical and research applications.

• Publisher: Cold Spring Harbor Laboratory
• Subject: Bioinformatics