Related Experiment Video
Updated: Feb 23, 2026
12:09
Network Analysis of the Default Mode Network Using Functional Connectivity MRI in Temporal Lobe Epilepsy
Published on: August 5, 2014
18.6K
Relationship between neuronal network architecture and naming performance in temporal lobe epilepsy: A connectome
B C Munsell1, G Wu2, J Fridriksson3
1College of Charleston, Department of Computer Science, Charleston, SC, USA.
Brain and Language
|September 14, 2017
Summary
Temporal lobe epilepsy (TLE) impairs naming by disrupting integrated brain networks. Machine learning identified key temporal lobe sub-networks crucial for naming, highlighting network integration over isolated areas.
Area of Science:
- Neuroscience
- Neurology
- Computational Psychiatry
Background:
- Impaired confrontation naming is a frequent symptom in temporal lobe epilepsy (TLE).
- The underlying neurobiological mechanisms and specific brain networks involved in naming deficits in TLE remain poorly understood.
- Understanding these networks can offer insights into naming processes across various neurological disorders.
Purpose of the Study:
- To investigate the structural brain networks supporting naming ability in individuals with medication-refractory TLE.
- To utilize a machine learning approach to predict naming performance based on brain connectome data.
Main Methods:
- A connectome-based prediction framework was employed using diffusion tensor imaging (DTI) to reconstruct structural brain connectomes.
- Network properties, specifically nodal eigenvector centrality (a measure of regional network integration), were extracted from anatomically defined brain regions.
- A multi-task machine learning algorithm coupled with support vector regression was used to predict naming performance.
Main Results:
- Nodal eigenvector centrality successfully predicted approximately 60% of the variance in naming performance.
- Key brain regions with the highest predictive weight were bilaterally distributed within perilimbic sub-networks, primarily involving medial and lateral temporal lobe areas.
- These findings suggest that intact naming relies on the integration of distributed sub-networks rather than isolated brain regions.
Conclusions:
- Naming in TLE is dependent on the integrated function of specific sub-networks, particularly those involved in semantic integration and lexical retrieval within the temporal lobes.
- Disruption of these integrated networks, rather than isolated areas, underlies naming impairments in TLE.
- This research contributes to understanding the large-scale structural network basis of language processing in neurological conditions.
