Document Type


Publication Date



This article is the author’s final published version in Frontiers in Human Neuroscience, Volume 15, October 2021, Article number 729677.

The published version is available at Copyright © Wu et al.


Background: Neuronal loss in Parkinson's Disease (PD) leads to widespread neural network dysfunction. While graph theory allows for analysis of whole brain networks, patterns of functional connectivity (FC) associated with motor response to deep brain stimulation of the subthalamic nucleus (STN-DBS) have yet to be explored.

Objective/Hypothesis: To investigate the distributed network properties associated with STN-DBS in patients with advanced PD.

Methods: Eighteen patients underwent 3-Tesla resting state functional MRI (rs-fMRI) prior to STN-DBS. Improvement in UPDRS-III scores following STN-DBS were assessed 1 year after implantation. Independent component analysis (ICA) was applied to extract spatially independent components (ICs) from the rs-fMRI. FC between ICs was calculated across the entire time series and for dynamic brain states. Graph theory analysis was performed to investigate whole brain network topography in static and dynamic states.

Results: Dynamic analysis identified two unique brain states: a relative hypoconnected state and a relative hyperconnected state. Time spent in a state, dwell time, and number of transitions were not correlated with DBS response. There were no significant FC findings, but graph theory analysis demonstrated significant relationships with STN-DBS response only during the hypoconnected state - STN-DBS was negatively correlated with network assortativity.

Conclusion: Given the widespread effects of dopamine depletion in PD, analysis of whole brain networks is critical to our understanding of the pathophysiology of this disease. Only by leveraging graph theoretical analysis of dynamic FC were we able to isolate a hypoconnected brain state that contained distinct network properties associated with the clinical effects of STN-DBS.

Creative Commons License

Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.

PubMed ID