Space-Time Modeling Group 

Clustering of Brain Signals

 

• In this project we develop a method that identifies brain regions that are synchronized during resting state in a sense that these regions share similar oscillations or waveforms. Our proposed method produces clusters of EEGs which serve as a proxy for segmenting the brain cortical surface. 

  We develop a new time series clustering method, which uses the total variation distance as a measure of similarity and the hierarchical merger algorithm as the clustering algorithm. 

• Researchers: C. Euan  (Lead, CIMAT), H. Ombao (UC Irvine) and J. Ortega (CIMAT)

• Collaborators: S. Cramer (Neurology, UC Irvine) and D. Moorman (Psychology, UMass).