Automatic parcellation of resting-state cortical dynamics by iterative community detection and similarity measurements <i>Running title: Functional parcellation of the cerebral cortex</i>
To investigate the properties of a large-scale brain network, it is a common practice to reduce the dimension of resting state functional magnetic resonance imaging (rs-fMRI) data to tens to hundreds of nodes. This study presents an analytic streamline that incorporates modular analysis and similari...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
AIMS Press
2021-09-01
|
| Series: | AIMS Neuroscience |
| Subjects: | |
| Online Access: | https://www.aimspress.com/article/doi/10.3934/Neuroscience.2021028?viewType=HTML |
| _version_ | 1818937266555322368 |
|---|---|
| author | Tien-Wen Lee Gerald Tramontano |
| author_facet | Tien-Wen Lee Gerald Tramontano |
| author_sort | Tien-Wen Lee |
| collection | DOAJ |
| description | To investigate the properties of a large-scale brain network, it is a common practice to reduce the dimension of resting state functional magnetic resonance imaging (rs-fMRI) data to tens to hundreds of nodes. This study presents an analytic streamline that incorporates modular analysis and similarity measurements (MOSI) to fulfill functional parcellation (FP) of the cortex. MOSI is carried out by iteratively dividing a module into sub-modules (via the Louvain community detection method) and unifying similar neighboring sub-modules into a new module (adjacent sub-modules with a similarity index <0.05) until the brain modular structures of successive runs become constant. By adjusting the gamma value, a parameter in the Louvain algorithm, MOSI may segment the cortex with different resolutions. rs-fMRI scans of 33 healthy subjects were selected from the dataset of the Rockland sample. MOSI was applied to the rs-fMRI data after standardized pre-processing steps. The results indicate that the parcellated modules by MOSI are more homogeneous in content. After reducing the grouped voxels to representative neural nodes, the network structures were explored. The resultant network components were comparable with previous reports. The validity of MOSI in achieving data reduction has been confirmed. MOSI may provide a novel starting point for further investigation of the network properties of rs-fMRI data. Potential applications of MOSI are discussed. |
| first_indexed | 2024-12-20T05:49:14Z |
| format | Article |
| id | doaj.art-bdd7afa050584ae899b3ceb49f992a27 |
| institution | Directory Open Access Journal |
| issn | 2373-7972 |
| language | English |
| last_indexed | 2024-12-20T05:49:14Z |
| publishDate | 2021-09-01 |
| publisher | AIMS Press |
| record_format | Article |
| series | AIMS Neuroscience |
| spelling | doaj.art-bdd7afa050584ae899b3ceb49f992a272022-12-21T19:51:13ZengAIMS PressAIMS Neuroscience2373-79722021-09-018452684210.3934/Neuroscience.2021028Automatic parcellation of resting-state cortical dynamics by iterative community detection and similarity measurements <i>Running title: Functional parcellation of the cerebral cortex</i>Tien-Wen Lee 0Gerald Tramontano 11. The Neuro Cognitive Institute (NCI) Clinical Research Foundation, NJ 07856, US 2. Department of Psychiatry, Dajia Lee's General Hospital, Lee's Medical Corporation, Taichung 43748, Taiwan1. The Neuro Cognitive Institute (NCI) Clinical Research Foundation, NJ 07856, USTo investigate the properties of a large-scale brain network, it is a common practice to reduce the dimension of resting state functional magnetic resonance imaging (rs-fMRI) data to tens to hundreds of nodes. This study presents an analytic streamline that incorporates modular analysis and similarity measurements (MOSI) to fulfill functional parcellation (FP) of the cortex. MOSI is carried out by iteratively dividing a module into sub-modules (via the Louvain community detection method) and unifying similar neighboring sub-modules into a new module (adjacent sub-modules with a similarity index <0.05) until the brain modular structures of successive runs become constant. By adjusting the gamma value, a parameter in the Louvain algorithm, MOSI may segment the cortex with different resolutions. rs-fMRI scans of 33 healthy subjects were selected from the dataset of the Rockland sample. MOSI was applied to the rs-fMRI data after standardized pre-processing steps. The results indicate that the parcellated modules by MOSI are more homogeneous in content. After reducing the grouped voxels to representative neural nodes, the network structures were explored. The resultant network components were comparable with previous reports. The validity of MOSI in achieving data reduction has been confirmed. MOSI may provide a novel starting point for further investigation of the network properties of rs-fMRI data. Potential applications of MOSI are discussed.https://www.aimspress.com/article/doi/10.3934/Neuroscience.2021028?viewType=HTMLcortexfunctional connectivityfunctional magnetic resonance imaging (fmri)functional parcellationresting state fmri (rs-fmri) |
| spellingShingle | Tien-Wen Lee Gerald Tramontano Automatic parcellation of resting-state cortical dynamics by iterative community detection and similarity measurements <i>Running title: Functional parcellation of the cerebral cortex</i> AIMS Neuroscience cortex functional connectivity functional magnetic resonance imaging (fmri) functional parcellation resting state fmri (rs-fmri) |
| title | Automatic parcellation of resting-state cortical dynamics by iterative community detection and similarity measurements <i>Running title: Functional parcellation of the cerebral cortex</i> |
| title_full | Automatic parcellation of resting-state cortical dynamics by iterative community detection and similarity measurements <i>Running title: Functional parcellation of the cerebral cortex</i> |
| title_fullStr | Automatic parcellation of resting-state cortical dynamics by iterative community detection and similarity measurements <i>Running title: Functional parcellation of the cerebral cortex</i> |
| title_full_unstemmed | Automatic parcellation of resting-state cortical dynamics by iterative community detection and similarity measurements <i>Running title: Functional parcellation of the cerebral cortex</i> |
| title_short | Automatic parcellation of resting-state cortical dynamics by iterative community detection and similarity measurements <i>Running title: Functional parcellation of the cerebral cortex</i> |
| title_sort | automatic parcellation of resting state cortical dynamics by iterative community detection and similarity measurements i running title functional parcellation of the cerebral cortex i |
| topic | cortex functional connectivity functional magnetic resonance imaging (fmri) functional parcellation resting state fmri (rs-fmri) |
| url | https://www.aimspress.com/article/doi/10.3934/Neuroscience.2021028?viewType=HTML |
| work_keys_str_mv | AT tienwenlee automaticparcellationofrestingstatecorticaldynamicsbyiterativecommunitydetectionandsimilaritymeasurementsirunningtitlefunctionalparcellationofthecerebralcortexi AT geraldtramontano automaticparcellationofrestingstatecorticaldynamicsbyiterativecommunitydetectionandsimilaritymeasurementsirunningtitlefunctionalparcellationofthecerebralcortexi |