Advances in functional magnetic resonance imaging of the human brainstem

Highlights

• Functional MRI of the brainstem is hindered by high physiological noise in this area.

• We introduce masked ICA as a means for spatial suppression of physiological noise.

• Brainstem resting-state networks are presented for the first time.

• Cortical connectivity and specificity were assessed using back-projection.

• 29 resting-state components were both reproducible and specific.

Abstract

The brainstem is of tremendous importance for our daily survival, and yet the functional relationships between various nuclei, their projection targets, and afferent regulatory areas remain poorly characterized. The main reason for this lies in the sub-optimal performance of standard neuroimaging methods in this area. In particular, fMRI signals are much harder to detect in the brainstem region compared to cortical areas. Here we describe and validate a new approach to measure activation of brainstem nuclei in humans using standard fMRI sequences and widely available tools for statistical image processing. By spatially restricting an independent component analysis to an anatomically defined brainstem mask, we excluded those areas from the analysis that were strongly affected by physiological noise. This allowed us to identify for the first time intrinsic connectivity networks in the human brainstem and to map brainstem–cortical connectivity purely based on functionally defined regions of interest.