Graph Regularized EEG Source Imaging with In-Class Consistency and Out-Class DiscriminationMS4

Electroencephalography (EEG) has become one of the most popular brain imaging tools. We develop a graph Laplacian-based model for discriminative EEG brain source imaging that takes into account the label information (happiness, sadness, surprise, etc). In addition, we also incorporate some regularization terms, such as L1 norm, total variation, and nuclear norm (low rank) to further improve the results. Simulated experiments show the effectiveness of the proposed framework.

This presentation is part of Minisymposium “MS4 - Graph Techniques for Image Processing (2 parts)
organized by: Yifei Lou (University of Texas at Dallas) , Jing Qin (Montana State University) .

Yifei Lou (University of Texas at Dallas)
image reconstruction, inverse problems