Gibson, William S.

OBJECTIVE Deep brain stimulation (DBS) of the subthalamic nucleus (STN) is an established neurosurgical treatment for the motor symptoms of Parkinson’s disease (PD). While often highly effective, DBS does not always yield optimal therapeutic outcomes, and stimulation-induced adverse effects, including paresthesia, muscle contractions, and nausea/lightheadedness, commonly occur and can limit the efficacy of stimulation. Currently, objective metrics do not exist for monitoring neural changes associated with stimulation-induced therapeutic and adverse effects. METHODS In the present study, the authors combined intraoperative functional MRI (fMRI) with STN DBS in 20 patients with PD to test the hypothesis that stimulation-induced blood oxygen level–dependent signals contained predictive information concerning the therapeutic and adverse effects of stimulation. RESULTS As expected, DBS resulted in blood oxygen level–dependent activation in myriad motor regions, including the primary motor cortex, caudate, putamen, thalamus, midbrain, and cerebellum. Across the patients, DBS-induced improvements in contralateral Unified Parkinson’s Disease Rating Scale tremor subscores correlated with activation of thalamic, brainstem, and cerebellar regions. In addition, improvements in rigidity and bradykinesia subscores correlated with activation of the primary motor cortex. Finally, activation of specific sensorimotor-related subregions correlated with the presence of DBS-induced adverse effects, including paresthesia and nausea (cerebellar cortex, sensorimotor cortex) and unwanted muscle contractions (caudate and putamen). CONCLUSIONS These results suggest that DBS-induced activation patterns revealed by fMRI contain predictive information with respect to the therapeutic and adverse effects of DBS. The use of fMRI in combination with DBS therefore may hold translational potential to guide and improve clinical stimulator optimization in patients.

AbstractBased on cytoarchitecture, the posterior cingulate cortex (PCC) is thought to be comprised of two distinct functional subregions: the dorsal and ventral PCC (dPCC and vPCC). However, functional subregions do not completely match anatomical boundaries in the human brain. To understand the relationship between the functional organization of regions and anatomical features, it is necessary to apply parcellation algorithms based on functional properties. We therefore defined functionally informed subregions in the human PCC by parcellation of regions with similar patterns of functional connectivity in the resting brain. We used various patterns of functional connectivity, namely local, whole‐brain and diffuse functional connections of the PCC, and various clustering methods, namely hierarchical, spectral, andk‐means clustering to investigate the subregions of the PCC. Overall, the approximate anatomical boundaries and predicted functional regions were highly overlapped to each other. Using hierarchical clustering, the PCC could be clearly separated into two anatomical subregions, namely the dPCC and vPCC, and further divided into four subregions segregated by local functional connectivity patterns. We show that the PCC could be separated into two (dPCC and vPCC) or four subregions based on local functional connections and hierarchical clustering, and that subregions of PCC display differential global functional connectivity, particularly along the dorsal‐ventral axis. These results suggest that differences in functional connectivity between dPCC and vPCC may be due to differences in local connectivity between these functionally hierarchical subregions of the PCC.Hum Brain Mapp 38:2808–2818, 2017. ©2017 Wiley Periodicals, Inc.