Limits...
BASCO: a toolbox for task-related functional connectivity.

Göttlich M, Beyer F, Krämer UM - Front Syst Neurosci (2015)

Bottom Line: BASCO supports seed-based functional connectivity as well as brain network analyses.Thus, BASCO allows investigating task-specific rather than resting-state networks.Here, we summarize the main features of the toolbox and describe the methods and algorithms.

View Article: PubMed Central - PubMed

Affiliation: Department of Neurology, University of Lübeck Lübeck, Germany.

ABSTRACT
BASCO (BetA Series COrrelation) is a user-friendly MATLAB toolbox with a graphical user interface (GUI) which allows investigating functional connectivity in event-related functional magnetic resonance imaging (fMRI) data. Connectivity analyses extend and compliment univariate activation analyses since the actual interaction between brain regions involved in a task can be explored. BASCO supports seed-based functional connectivity as well as brain network analyses. Although there are a multitude of advanced toolboxes for investigating resting-state functional connectivity, BASCO is the first toolbox for evaluating task-related whole-brain functional connectivity employing a large number of network nodes. Thus, BASCO allows investigating task-specific rather than resting-state networks. Here, we summarize the main features of the toolbox and describe the methods and algorithms.

No MeSH data available.


ROI-based network analysis (single subject level). (A) Single subject network matrix using the AAL parcellation (Tzourio-Mazoyer et al., 2002). An absolute threshold of w > 0.7 was applied to binarize the matrix. (B) Distribution of Fisher z-transformed correlation coefficients. (C) AAL atlas ROIs showing a high functional connectivity to the precuneus. (D) Regions showing a high degree centrality, i.e., one standard deviation above the mean (indicated by the solid and dashed lines).
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4565057&req=5

Figure 3: ROI-based network analysis (single subject level). (A) Single subject network matrix using the AAL parcellation (Tzourio-Mazoyer et al., 2002). An absolute threshold of w > 0.7 was applied to binarize the matrix. (B) Distribution of Fisher z-transformed correlation coefficients. (C) AAL atlas ROIs showing a high functional connectivity to the precuneus. (D) Regions showing a high degree centrality, i.e., one standard deviation above the mean (indicated by the solid and dashed lines).

Mentions: In order to illustrate the ROI-based network analyses which can be performed using BASCO, we applied a brain parcellation according to the AAL atlas (excluding the cerebellum) to the data and extracted the mean beta-series for each region. We selected only trials and their corresponding beta-values which are related to emotional stimuli and calculated the network matrix correlating all 90 regional beta-series. This resulted in a 90 × 90 network matrix which is shown in Figure 3A (single subject). For the purpose of this demonstration the network matrix was thresholded keeping only weights, i.e., correlation coefficients, larger than w = 0.7. All remaining weights were set to one. This resulted in a binary, undirected connectivity matrix. Figure 3B shows all entries of the network matrix before a threshold was applied. Note, that the correlation coefficients were Fisher z-transformed. BASCO offers the functionality to conveniently visualize all brain regions which are strongly connected to a given seed-region. This is illustrated in Figure 3C. The left precuneus was chosen as a seed region. Evaluating the connectivity matrix, we found that the precuneus is connected to the left angular gyrus, left cingulate cortex and the medial frontal cortex (bilateral). The degree centrality is a simple metric for characterizing the role of individual nodes within a network. It refers to the number of connections to other nodes in the network. Nodes with a high degree centrality can be regarded as hubs important for network integration. Figure 3D shows the network nodes with a degree centrality one standard deviation above the mean.


BASCO: a toolbox for task-related functional connectivity.

Göttlich M, Beyer F, Krämer UM - Front Syst Neurosci (2015)

ROI-based network analysis (single subject level). (A) Single subject network matrix using the AAL parcellation (Tzourio-Mazoyer et al., 2002). An absolute threshold of w > 0.7 was applied to binarize the matrix. (B) Distribution of Fisher z-transformed correlation coefficients. (C) AAL atlas ROIs showing a high functional connectivity to the precuneus. (D) Regions showing a high degree centrality, i.e., one standard deviation above the mean (indicated by the solid and dashed lines).
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4565057&req=5

Figure 3: ROI-based network analysis (single subject level). (A) Single subject network matrix using the AAL parcellation (Tzourio-Mazoyer et al., 2002). An absolute threshold of w > 0.7 was applied to binarize the matrix. (B) Distribution of Fisher z-transformed correlation coefficients. (C) AAL atlas ROIs showing a high functional connectivity to the precuneus. (D) Regions showing a high degree centrality, i.e., one standard deviation above the mean (indicated by the solid and dashed lines).
Mentions: In order to illustrate the ROI-based network analyses which can be performed using BASCO, we applied a brain parcellation according to the AAL atlas (excluding the cerebellum) to the data and extracted the mean beta-series for each region. We selected only trials and their corresponding beta-values which are related to emotional stimuli and calculated the network matrix correlating all 90 regional beta-series. This resulted in a 90 × 90 network matrix which is shown in Figure 3A (single subject). For the purpose of this demonstration the network matrix was thresholded keeping only weights, i.e., correlation coefficients, larger than w = 0.7. All remaining weights were set to one. This resulted in a binary, undirected connectivity matrix. Figure 3B shows all entries of the network matrix before a threshold was applied. Note, that the correlation coefficients were Fisher z-transformed. BASCO offers the functionality to conveniently visualize all brain regions which are strongly connected to a given seed-region. This is illustrated in Figure 3C. The left precuneus was chosen as a seed region. Evaluating the connectivity matrix, we found that the precuneus is connected to the left angular gyrus, left cingulate cortex and the medial frontal cortex (bilateral). The degree centrality is a simple metric for characterizing the role of individual nodes within a network. It refers to the number of connections to other nodes in the network. Nodes with a high degree centrality can be regarded as hubs important for network integration. Figure 3D shows the network nodes with a degree centrality one standard deviation above the mean.

Bottom Line: BASCO supports seed-based functional connectivity as well as brain network analyses.Thus, BASCO allows investigating task-specific rather than resting-state networks.Here, we summarize the main features of the toolbox and describe the methods and algorithms.

View Article: PubMed Central - PubMed

Affiliation: Department of Neurology, University of Lübeck Lübeck, Germany.

ABSTRACT
BASCO (BetA Series COrrelation) is a user-friendly MATLAB toolbox with a graphical user interface (GUI) which allows investigating functional connectivity in event-related functional magnetic resonance imaging (fMRI) data. Connectivity analyses extend and compliment univariate activation analyses since the actual interaction between brain regions involved in a task can be explored. BASCO supports seed-based functional connectivity as well as brain network analyses. Although there are a multitude of advanced toolboxes for investigating resting-state functional connectivity, BASCO is the first toolbox for evaluating task-related whole-brain functional connectivity employing a large number of network nodes. Thus, BASCO allows investigating task-specific rather than resting-state networks. Here, we summarize the main features of the toolbox and describe the methods and algorithms.

No MeSH data available.