Stimulus profile and modeling of continuous galvanic vestibular stimulation in functional magnetic resonance imaging

Both direct current (DC) and alternating current (AC) have been used for galvanic vestibular stimulation (GVS) during functional MRI (fMRI). The perceptual effects of apparent rotation persist while the current is being applied, but there is a subjective decay over the entire stimulation period. Particularly during DC-GVS in a supine position, subjects report a strong vestibular sensation related to the onset and offset of the stimulus, and weaker effects during constant DC-GVS stimulation. In the present study, we analyzed DC-GVS fMRI data from a group of volunteers with two different statistical models. In model I, the effects of GVS were modeled as a single regressor that described the periods during which the current was switched on. In model II, an additional regressor describing the onset and offset of the stimulation was included. The activation pattern found by using model I included only a subset of the activation patterns known to respond to vestibular stimulation from previous imaging studies. Model II revealed two different activation maps: block effects similar to the results obtained in model I and additional transient GVS effects with larger activation clusters and higher t-values. This extended activation pattern resembled the results obtained during AC-GVS, including multisensory vestibular projection areas. We show that the major part of the blood oxygenation level dependent (BOLD) responses elicited by DC-GVS is related to the on- and offset of stimulation currents and reflects the perceptual experience just-described. In addition, the separate modeling of transient and persistent effects of DC-GVS can identify distinct cortical correlates of those effects.