Altered Macroscale Topography of Gamma Oscillations in Parkinson's Disease

1635 

Abstract Submission 

Hao Ding¹, Jens Volkmann², Muthuraman Muthuraman³

¹Würzburg, wurzburg, bavaria, ²Universitätsklinikum Würzburg, Würzburg, Bavaria, ³Universitätsklinikum Würzburg, Würzburg, Germany

Hao Ding  
Würzburg
wurzburg, bavaria

Jens Volkmann  
Universitätsklinikum Würzburg
Würzburg, Bavaria

Muthuraman Muthuraman  
Universitätsklinikum Würzburg
Würzburg, Germany

In Parkinson's disease (PD), reduced gamma oscillations disrupt the functional connectivity between brain regions, affecting macroscale brain function. Entraining the gamma rhythm through neural modulation can help restore the cortical plasticity related to motor function. However, the macroscale functional topography at specific neural oscillations is not well characterized. This study aims to examine the macroscale functional reorganization in PD, with a particular focus on the gradient topography associated with gamma oscillations.

Thirty-five PD patients (in the medication-On state) and 35 sex and age-matched healthy seniors were studied. A 5-minute resting state was recorded using high-density EEG. The forward solution was estimated using the finite element method, and cortical brain activity was reconstructed using a beamformer as the inverse solution. Functional connectivity in the source space was calculated using Pearson correlation based on HCP-MMP1 parcellation. Cortical gradients were analysed using a normalized angle kernel to capture connectivity profile similarities. Primary gradient components at gamma band (60-90Hz) were identified using diffusion map embedding. Network-wise analysis was performed on the Cole-Anticevic brain-wide network parcellation.A surface-based linear model on z-transformed gradients was used to identify gradient differences between groups. Significant clusters were functionally decoded using the Neurosynth meta-database. Finally, clinical measures were correlated with these clusters to provide meaningful clinical interpretations.

An examination of the primary gradient components revealed alterations in gamma oscillation in PD. Further exploration through Euclidean space analysis identified an additional cluster in PD patients compared to healthy individuals. Notably, global histograms also showed a noticeable contraction of the two gradient anchors in PD patients. A network-wise analysis highlighted increased gradient values in the Cingulo-Opercular network (p = 0.032). Surface-based linear models revealed differences on the first gradient component in the middle temporal cortex (p=0.018). Importantly, gradients in this region showed a significant negative correlation with UPDRS-III scores (p=0.03), highlighting their clinical importance. Functional annotation of brain regions suggested their significance in the motor network, as interpreted from meta-analytic maps.

In conclusion, our investigation has revealed disruptions in the macroscale hierarchy, which affect the processes of integration and segregation within unimodal and transmodal networks in PD patients. These findings provide valuable insights into the subtle changes associated with PD at both macroscopic and clinical levels. Future research on the topography of gradients in other canonical neural oscillations could provide deeper insights into the pathology of PD.

EEG/MEG Modeling and Analysis ¹

EEG

Neurophysiology of Imaging Signals ²

Computational Neuroscience

Cortex

Electroencephaolography (EEG)