Rootlets-informed registration to a spinal cord template: proof-of-concept

Motivation: Spinal cord fMRI studies require precise localization of spinal levels for reliable group analyses. Traditional template-based registration using intervertebral discs is imprecise due to anatomical variability. Goal(s): To improve spinal cord MRI template registration accuracy by incorporating nerve rootlet information, enhancing normalization and reproducibility across subjects. Approach: We developed a registration method leveraging dorsal cervical rootlets segmentation and aligning them non-linearly with the PAM50 spinal cord template. Validation was performed on a multi-subject dataset and single-subject dataset with various neck positions. Results: Rootlet-based registration showed superior alignment accuracy and consistency across subjects compared to traditional disc-based methods, promising enhanced fMRI analysis reliability. Impact: Incorporating nerve rootlets into spinal cord MRI template registration improves the alignment of spinal levels, enhancing the accuracy and reproducibility of group analyses in fMRI studies. This advancement allows more precise mapping of the spinal cord across individuals.