Motivation: Resting-state functional MRI (rs-fMRI) could aid neurosurgical planning for patients unable to perform task-based (tb-)fMRI tasks, however, complex analyses and lack of validated seeds for language mapping hinder clinical adoption. Goal(s): Develop a clinically feasible, automated rs-fMRI pipeline for language and motor mapping, optimised for comparability to tb-fMRI. Approach: We developed a pulse sequence, automated preprocessing and analysis pipeline, with standardised seeds. Results from 10 healthy volunteers were compared to atlases and tb-fMRI. Results: Language and motor resting-state networks (RSNs) showed high sensitivity and high/moderate specificity compared to atlases. Optimal RSN thresholds for comparability to tb-fMRI were identified, producing moderate overlap with tb-fMRI. Impact: The automated resting-state functional MRI (rs-fMRI) pipeline, using standardised seeds, produces resting-state networks optimised for comparison with task-based (tb)-fMRI and shows promise as an alternative for preoperative language and motor mapping. Future work will explore language lateralisation and rs-fMRI-driven tractography.
Verdon et al. (Tue,) studied this question.