Related Experiment Video
Updated: May 1, 2026
08:51
Statistical Modelling of Cortical Connectivity Using Non-invasive Electroencephalograms
Published on: November 1, 2019
6.1K
Untamed: Unconstrained Tensor Decomposition and Graph Node Embedding for Cortical Parcellation
Yijun Liu1, Jian Li2,3, Jessica L Wisnowski4,5
1Ming Hsieh Department of Electrical and Computer Engineering, University of Southern California, Los Angeles, California, USA.
Human Brain Mapping
|March 6, 2026
Summary
Untamed, a new framework, uses tensor decomposition and graph embedding for brain parcellation. It creates spatially coherent regions aligned with functional networks, improving analysis of neuroimaging data.
Area of Science:
- Neuroscience
- Neuroimaging
- Computational Biology
Background:
- Cortical parcellation is crucial for interpreting neuroimaging data.
- Optimal spatial features for parcellation from resting-state fMRI (rsfMRI) are unclear.
- Existing methods like Independent Component Analysis (ICA) have limitations.
Purpose of the Study:
- To introduce Untamed, a novel framework for cortical parcellation.
- To integrate unconstrained tensor decomposition and graph node embedding.
- To generate spatially coherent cortical regions aligned with functional networks.
Main Methods:
- Utilized unconstrained tensor decomposition (NASCAR) for functional network identification.
- Employed graph node embedding for generating cortical parcellations.
- Developed an automated pipeline for rapid adaptation and custom parcellation generation.
Main Results:
- Untamed produces near-homogeneous, spatially coherent regions.
- Regions align well with large-scale functional networks.
- Demonstrated improved or comparable performance in functional connectivity homogeneity and task contrast alignment.
Conclusions:
- Untamed offers an effective approach to cortical parcellation using rsfMRI data.
- The framework avoids strong assumptions of methods like ICA.
- Publicly available atlases and code facilitate further research in brain mapping.
