Deep brain stimulation (DBS) through implantable neurostimulation electrodes that affect dopaminergic control is an important symptomatic therapy in movement disorders, and has been shown to affect reinforcement learning and motivation1. Despite its success, the neurophysiological mechanisms underlying the efficacy of DBS remain a subject of debate. Performed concurrently with DBS, PET/MR functional imaging can track dopaminergic output together with functional activation and thus integrate local neurochemical modulation with function simultaneously. This approach has the potential to shed light on the underlying molecular mechanisms of DBS for the evaluation of existing and novel brain targets.A microstimulation DBS electrode was implanted unilaterally in the right ventral tegmental area (VTA) of a non-human primate (NHP, baboon), as shown in Figure 1. Following recovery from surgery, the animal underwent 7 imaging sessions with simultaneous PET/MRI during electrode stimulation. Stimulation was performed with 200 ms pulses at 100 Hz, applied every 8 s for 0.95 min at 2 microelectrodes (out of a bundle of 23 implanted microwires). This ON stimulation period was interleaved with 0.85 min of rest and repeated for a total of 10 min at a time. Variations of stimulation patterns were carried out in order to determine their effects on imaging signals. The radiotracer [11C]raclopride was administered as a bolus + infusion to detect within-scan changes in dopamine at D2/D3 receptors due to stimulation. Dynamic PET data were acquired and analyzed with the simplified reference tissue kinetic model2, in which the binding parameter was allowed to change during stimulation blocks. Occupancy values were calculated as the percent change from baseline binding potential (BPND). At the start of fMRI acquisition with gradient-echo echo-planar-imaging, iron oxide (Feraheme) was administered to increase detection power. fMRI data were analyzed with the GLM. Data from multiple imaging sessions were analyzed as a group using a mixed effects model.fMRI results showed a robust positive CBV change (corresponding to negative signal changes shown in Figure 2) in the nucleus accumbens on the side of the DBS electrode implant (right side), both in single sessions as well as in the group analysis. Average signal magnitudes were 2.3% ± 0.3% for the activated region seen in Figure 2. Alterations in stimulation patterns affected fMRI magnitude, with longer, continuous stimulations of 3x10 min showing a maximum magnitude of 4.12% ± 0.4% and stimulation patterns of 10 min ON and OFF blocks showed a maximum magnitude of 2.67% ± 0.3%. In the group analysis, BPND decreases from baseline were observed during stimulation, in part of the right putamen (Figure 2) with an average value of 0.21 ± 0.14. Baseline BPND in the same region were 4.4 ± 0.5 on average. Changes in BPND translate to an average receptor occupancy of 4.6% ± 2.5%. Figure 3 shows a representative time-activity curve from a single session, during which stimulation occurred in four blocks of 10 min, with 10 min rest in between. Subtle changes in the timecourse can be seen in the ROI that corresponds to the area of change in Figure 2 on the side with the DBS implant but less so on the contralateral side.This study shows the feasibility of doing concurrent DBS with combined PET and fMRI acquisitions. The group analysis shows that signal changes from fMRI and [11C]raclopride-PET occur in similar regions of the basal ganglia, along dopaminergic projections from the VTA. While fMRI signal changes are significant at a single session level, PET signal changes due to stimulation are much smaller and their magnitudes are within the limit of detectability, given the noise in the time activity curves for small regions. Given observed fMRI signal change magnitudes of 2.3%, occupancy changes as predicted by a neurovascular coupling model3 are expected to be small (6%), which is within the range of our results. If dopamine at D2/D3 receptors is driving the fMRI response, we would expect to see negative CBV signal changes due to agonist binding at D2/D33. It is thus especially interesting that our results show a consistent positive CBV change. This suggests that dopamine at postsynaptic D2/D3 receptors is not the only driving force behind the fMRI signal. Rather, dopamine binding to D1 receptors or non-dopaminergic contributions may also play a role4,5. Further exploration of DBS parameter space, including optimization of electrode selection, applied current, and stimulation paradigm, can start to clarify the magnitude of dopamine efflux in the basal ganglia due to DBS of the VTA and the relationship of dopamine efflux to fMRI signal in this paradigm.