The overall goal of this analysis is to identify memory-related context-dependent changes to functional connectivity between regions of the hippocampus and the rest of the brain. This method can help answer key questions in the field of cognitive neuroscience such as how does functional connectivity change in response to specific cognitive demands within an experimental task. The main advantage of this technique is that it allows researchers to test specific hypotheses regarding the functional connectivity of key brain regions during a cognitive task.
For this experiment, include individual aged 55 and older with cognitive decline who have been genotyped for the Alzheimer's disease risk allele apolipoprotein E epsilon four prior to the experiment, screen subjects for MRI safety and obtain informed consent. Use a three-tesla MRI system to acquire whole brain imaging data. For functional imaging, collect axial slices using an echo-planar imaging sequence while running an unrelated words-associative memory task.
To facilitate registration of the functional images, also acquire axial slices of T2-weighted coplanar structural images. For high-resolution structural imaging, collect axial slices using a 3D T1-weighted sequence. Once imaging is complete for all participants, set up preprocessing steps and the first level general linear model using FSL FMRI expert analysis tool, or FEAT, for the first participant.
In the Data tab, click on Select 4D data and navigate to the motion corrected and brain extracted file. Set the TR to match that of the functional sequence and use the default high pass filter. Now, in the Pre-stats tab, click none under motion correction and uncheck BET brain extraction.
Enter five millimeters to set the full width half maximum Gaussian kernel for spatial smoothing. Then, click Full model setup and create the task timing files denoting the onset and offset of the task phases. Add these to the GLM by choosing three-column format and navigating to the relevant text file.
Include one for the encoding phase of the task and one for the retrieval phase. For convolution, choose the Double-Gamma HRF option. Next, use the output of the MCFLIRT tool to create six single-column text files that describe the motion correction performed at each volume within the data set.
Select Full model setup and add the parameters and their temporal derivatives as explanatory variables, or EVs, in the GLM. For each motion EV, choose Custom for basic shape, none for convolution and temporal filling. Now, navigate to the Stats tab in the software and select the output of the FSL motion outliers tool under the Add additional confound EVs option.
Now, in the Registration tab, check Expanded functional image and Main structural image for a two-step registration. Select the participants coplanar T2-weighted structural scan for the first step to register the functional to the structural data. Choose six degrees of freedom in the second dropdown box.
For the next step, register the T2-weighted image to the high-resolution T1-weighted MP-RAGE by selecting boundary based registration from the dropdown box. Finally, register the high-resolution structural data to the standard MNI 152 template selecting 12 degrees of freedom and a linear transformation. Before setting up the psychophysiological interaction model, first load the preprocess data in FSL FEAT software.
Choose the denoised image as the input file. In the Pre-stats tabs, set motion correction and brain extraction to None. Do not perform temporal filtering or spatial smoothing.
Then, in the Stats tab, select Full model setup and in the EVs tab, add all the variables from the first-level modeling including motion correction, the confound matrix from FSL motion outliers and task timing. Include an EV for the physiological time course from the seed as a covariate of no interest. Next, create the PPI terms by choosing Interaction in the Basic shape menu, and select the seed time course EV and one task EV.For the Make zero option, choose Center for the task variables and Mean for the seed time course EV.Now, in the Contrasts and F-tests tab, model the following specific effects by entering one in the corresponding EV cells.
Encoding task phase, retrieval task phase, seed time course, PPI of seed and encoding and PPI of seed and retrieval. Lastly, enter negative one to model negative PPIs for each task phase. Use statistical parametric mapping software tools to run group level comparisons.
Begin by selecting Specify second level, then select Two-sample T-test under Design. Navigate to the directory with the parameter estimate images for group one and select them. Then, add the images for group two and run this comparison by clicking on the Play button.
Now, return to the main window. Select Estimate and navigate to the SPM. mat file created in the previous step to run the model estimation.
Next, under the Results tab, select Define a new contrast. Choose T-contrast and enter one negative one in the Contrast box for APOE-4 carriers greater than APOE-4 non-carriers, then click OK.Finally, run group comparison contrasts as seen here. Choose None for Apply masking, and then manually set the voxel-level threshold and the cluster size minimum according to output from AFNI's 3dClustSim software.
Enter negative one-one for APOE-4 non-carriers greater than APOE-4 carriers. Within group generalized psychophysiological interaction analyses revealed significant decreases in functional connectivity in APOE-4 carriers, green, for both task conditions and hippocampal sub-regions. In APOE-4 non-carriers, red, significant decreases in functional connectivity were only observed with posterior hippocampus during encoding.
During retrievals, significant differences between APOE-4 carriers and non-carriers were found in the left supramarginal gyrus, dark blue, the right supramarginal angular junction, orange, as well in the right precuneus, purple. The peak coordinate for each cluster is reported in MNI space. Here, contrasts of parameter estimates from each cluster are plotted by group.
The red lines indicate zero and highlight that carriers have decreased functional connectivity to anterior hippocampus in these regions during retrieval. The band within the boxes represents the median, while the upper and lower edges of the boxes represent the first and third quartiles respectively. After its development, this technique paved the way for functional neuroimagers to explore dynamic task-related connectivity in humans.
This includes both healthy and patient cohorts as well as individuals at increased genetic risk for disease as we describe here. After watching this video, you should have a good understanding of how to use a PPI analysis to test for context-dependent functional connectivity changes between your seed region of interest and the rest of the brain.
