Abstract #3804

Morphological component analysis of functional MRI data based on sparse representations and dictionary learning

Hien M. Nguyen¹, Jingyuan Chen², and Gary H. Glover³

¹Department of Electrical Engineering & Information Technology, Vietnamese - German University, Binh Duong New City, Vietnam, ²Department of Electrical Engineering, Stanford University, Stanford, CA, United States, ³Department of Radiology, Stanford University, Stanford, CA, United States

A data-driven method for identifying functional connectivity networks utilizing sparse representations is presented. Specifically, fMRI signals are decomposed into morphological components which have sparse spatial overlap. Allowing sparse spatial overlap between components is more physically plausible than the statistical independence assumption of the Independent Component Analysis (ICA) method. The proposed formulation is related to the Morphological Component Analysis (MCA) and uses a K-Singular Value Decomposition (SVD) algorithm for dictionary learning. Experimental results prove that the MCA-KSVD method can identify functional networks in task and resting-state fMRI and thus can be used as an alternative method for investigating brain functional connectivity.

This abstract and the presentation materials are available to members only; a login is required.

Join Here