A Method for Investigating Age-related Differences in the Functional Connectivity of Cognitive Control Networks Associated with Dimensional Change Card Sort Performance

The overall goal of this procedure is to identify brain networks associated with task switching and test for age-related differences in the functional connectivity of these networks. This is accomplished by first collecting FMRI data from children and adults while completing the dimensional change card sort task. After pre-processing the data, functional volumes are decomposed into a set of statistically independent sources or components.

The components of relevance to executive functioning and task performance are selected by means of template matching and linear modeling respectively. The final step is to determine whether functional connectivity within the selected components differs for adults and children. Ultimately, ICA based FMRI analyses can be used to identify networks associated with task switching and test for age related differences in the functional organization of these networks.

The main advantage of this technique over existing techniques such as resting state functional connectivity analysis is it provides a means of accurately characterizing the function of cortical networks. This method can help answer key questions in the field of developmental cognitive neuroscience, such as whether broad scale cortical networks involved in high level cognitive operations such as tasks switching, undergo functional reorganization over development. To begin acquire FMRI data following procedures suitable for young children limit unwanted confounds by training participants in advance in a mock scanner facility when ready, implement a repeated trials version of the dimensional change card sort task.

Each run includes two eight trial switch blocks and two eight trial repeat blocks where switch blocks consist of four switch trials and four repeat trials and repeat blocks consist of eight repeat trials after data collection pre-processed the FMRI data following standard FMRI pre-processing procedures first download and install gif a group ICA software toolbox that works together with SPMA, well-known FMRI analysis package. Once downloaded, add the gift toolbox and all sub directories to the MATLAB search path and save the path file computing. A group ICA on FMRI data using gif makes substantial demands on RAM memory.

To avoid memory issues, it is best to run the ICA analysis on a server. If running the analysis on a local computer, the RAM requirements can be estimated using a script that is part of gif to parameterize the analysis, modify a preexisting batch script called Input data subjects one M that is stored in gif under ICAT B batch files. This can also be done by using gifs graphical user interface.

However, it is much easier with a bit of practice to set up the analysis by modifying this preexisting script. Next, specify the data modality. As FMRI specify the type of analysis as ICA to run the ICA with the I Casso procedure.

Select two under type of analysis and then parameterize the I Casso procedure in the succeeding lines of the setup file. Maximize the performance of the group PCA by choosing one under group PCA performance settings to enable sorting from a standard SPM design matrix specify whether or not there are different matrices for different subjects. Next, specify where the pre-processed functional data are stored and whether or not a SPM dot mat file containing the design matrix is stored together with the pre-processed functional data.

If each participant has the same number of runs, select one for method one. Under the data selection method option, complete the parameter source steer file pattern, flag, location by including the file path where the data are stored, the file format of the data and a statement indicating that individual sessions are stored as sub directories within each subject folder indicate the directory where the output of the analysis should be written. This should not be the same directory where the data are stored.

The next step is to provide a prefix that will be added to all output files. Next, provide a file path to a mask. At a minimum, the mask should eliminate skull extra cerebral space and especially the eyeballs as illustrated here.

Notice in this example residual evidence of the eyeballs. This will lead to a suboptimal ICA decomposition and should be avoided for the type of group PCA to be used. Choose subject specific to obtain the best time course.

Select GICA as the back reconstruction method. When selecting a data, pre-processing type, use intensity normalization to avoid non-numerical values in the output. In this example, the default of one is chosen.

Choose the standard PCA type and accept default values under PCA options. Next, specify how many PCAs to run on the data before the ICA. In addition, specify how many components to retain after each PCA use Z-score scaling to score the data, choose a blind source separation algorithm for the ICA for this work.

Infomax was used once the ICA is completed. Select components of potential theoretical interest for further consideration. Spatial sorting can be selected to sort components by means of spatial correlation with a preexisting template while temporal sorting sorts.

The component time courses by means of linear predictors from the SPM design matrix. Once the analysis is complete, test whether the child and adult versions of the selected components differ.Shown. Here are results of the ICAS O procedure applied to an ICA with 60 components.

Each numbered point represents one component. The black points represent different ICA decompositions of the same data with different random seeds. Ideally, results from different decompositions seen here as different black spots should cluster tightly around the numbered points.

This indicates good reliability in the decomposition. The executive control template and the selected right frontal parietal component are overlaid on identical slices from a high resolution anatomical scan and appear quite comparable. Functional connectivity was stronger in the medial prefrontal cortex and ventral segmental area and dorsal prefrontal and inferior parietal cortex of adults when compared to children.

Well, it's attempting this procedure. It's important to remember to check your output at each stage of the analysis. Small errors will propagate forward and lead to uninterpretable results.

Following this procedure. Other methods can be performed such as functional network connectivity analysis to answer additional questions such as whether functional interactions between components change as a function of task demands or participant age.