Regions of interest for resting-state fMRI analysis determined by inter-voxel cross-correlation.

In investigations of the brain's resting state using functional magnetic resonance imaging (fMRI), a seed-based approach is commonly used to identify brain regions that are functionally connected. The seed is typically identified based on anatomical landmarks, coordinates, or the location of brain activity during a separate task. However, anatomical boundaries may be difficult to discern, and designing a task to interrogate desired brain regions of interest may be difficult, especially when subject compliance is in question, as in many patient studies. In this study, a seed selection method based on inter-voxel cross-correlation of resting-state signals (i.e., a rest-based seed) is introduced. This method was used to determine resting-state connectivity between the left and right motor cortices in fifteen healthy right-handed subjects, and results were compared to seed selection based on the most significantly activated voxels during a separate task (i.e., a task-based seed). The z-coordinate of the centers of mass of the rest-based and task-based seeds within motor cortex were significantly different; task-based seeds were closer to the pial surface. Connectivity maps generated by rest-based seeds and task-based seeds were statistically equivalent; however, only 3 min of data were required to reach significance for rest-based seeds compared to an estimated 6 min for task-based seeds. Rest-based seeds also exhibited good inter-experimenter reproducibility. These findings suggest that seed regions based on inter-voxel cross-correlation of resting-state signals can be used as an alternative approach for connectivity analysis when task-related activity is not available or difficult to acquire, as in some patient studies.