使用功能性磁共振成像连接在颞叶癫痫默认模式网络的网络分析

The overall goal of the following experiment is to obtain statistical maps. Comparing the difference of resting state FMRI scans between healthy controls and patients with epilepsy. This is achieved by obtaining FMRI data in subjects with temporal lobe epilepsy, as well as healthy control subjects.

As a second step, the FMRI is pre-processed, which prepares the data to be statistically analyzed. Next, the pre-processed FMRI data is analyzed in order to obtain statistically valid comparisons between the two groups. Results are obtained that show differences in brain networks among patients with epilepsy compared to brain networks of healthy control subjects.

Based on statistical differences in brain network maps between these two groups, It's become increasingly apparent that even epilepsies that are manifest by focal seizures are emergent properties of diffuse network disturbances. The idea for this technique came after recognizing the importance of the DEF Favo network during seizures and wondering whether the default LO network is abnormal between seizures in temporal lobe epilepsy. The study population for this protocol should include three groups, right?

Temporal lobe epilepsy patients, left temporal lobe epilepsy patients, as well as healthy controls. A total of about 35 subjects are recommended. The epilepsy subject groups should be patients who are diagnosed to have temporal lobe epilepsy, such as candidates for anterior temporal lobe resection as determined by video, EEG monitoring, PET imaging and neuropsychological testing assure that all subjects have normal brain MRIs and are free from neurologic illness other than epilepsy in the patient groups.

Obtain IRB approval and written informed consent from all subjects prior to imaging and screen for MRI.Safety. Patients should continue their usual medications during the FMRI scan and should not be scanned immediately following a seizure. A three Tesla MRI system should be used for all imaging described in this protocol.

Obtain axial slices for functional images using an echo planar imaging sequence and for anatomic images, use a spoiled gradient recalled sequence. Ask participants to relax and remain still with their eyes closed, and perform functional imaging using the parameters as seen here. Also use the following parameters for the S PGR T one weighted high resolution structural imaging.

Each imaging session should last about 20 minutes. Begin by pre-processing the FMRI data using FSL software First. Use FSL MFL to remove head movement artifact.

Then use the FSL brain extraction tool or BET to remove non brainin tissue with the option dash F for bold files. This allows for further analysis steps on the brain tissue alone. Next in feet, run a minimally processed analysis with registration.

Select first level analysis and change full analysis to pre stats from the top two buttons. Then under pres stats tab, uncheck bet brain extraction, and select none for motion correction as these were already performed. Then register the functional images to the anatomical images, and then to a standard MNI image.

This results in the generation of transformation matrices, which are used later during analysis to warp the seed selected in standard space into the subject's brain space. Next, use the generated transformation matrix named standard to example funk dot mat and transform CSF and white matter ROIs into the individual bold space. Then using the FSL mean TS command, extract the time series from the CSF and white matter ROIs.

Using the ROI in individual subject space as a mask normalize the extracted time series using the software R.These time series will later be used as regressors in the general linear model to remove the corresponding artifactual signals from the analysis. The next step is removal of subject motion related artifacts. For regression of the motion parameters.

Set the following within FSL feet before running it first within the data tab, use the motion corrected and brain extracted file as inputs and set the TR value to correspond to your dataset. Set high pass filtering using a 102nd filter, which will remove very low frequency signals of no interest. A lowpass filter to remove high frequency signals will be applied later within the pres stats tab.

Choose none under motion correction and uncheck, but brain extraction. As these steps have been performed, perform spatial smoothing using a five millimeter full width half maximum. Then within the stats tab, regress the six motion parameters and their temporal derivatives.

Select none for convolution and check apply temporal filtering. Use the output of F selmic flirt to get text files of movement parameters, which can then be input into the feet analysis to regress these in a general linear model. Also, add the CSF and white matter signals that were extracted and normalized in previous steps to the GLM.

Select none for convolution. Add temporal derivative and uncheck apply temporal filtering. The residuals from pre-processing described above should be used for seed based correlation.

These residuals should be first passed through a low PESS filter of 0.1 hertz, then demeaned by subtracting the mean, dividing by the standard deviation and then scaled by adding 100 seed should be defined with a diameter of six millimeters. In the standard MNI space. Using MRI crown software, the posterior and the anterior seeds should correspond to the coordinates as seen here.

Note that these seed locations have been defined within healthy controls. The seeds should subsequently be transformed to each subject's individual functional brain space from the standard MNI space. For this, use the transformation matrix previously generated to transform the seed from standard m and i space to the individual functional space.

Next, use the FSL mean Ts command to extract the time series from the previously demeaned and scaled residual. Using the seed in the individual subject space as a mask. Normalize the extracted time series using the software R partial correlations between the seed voxels and all of their brain voxels should be calculated separately for each subject for each run.

For this, within the FSL feet interface, select first level analysis and then stats plus post stats within the data tab. The previously demeaned and scaled residual should be used as input. Set the high pass filter cutoff to 10, 000 as the residual is already high, passed at 100 seconds within the stats tab.

Unselect use film pre whitening and use the previously extracted and normalized seed time series. In the GLM within the post stats tab, set the desired Z stat threshold to a value of 2.0 prior to running group analysis. Combining runs within subjects, a Fisher Z transformation should be performed on the contrast of parameter estimates.

File generated from the correlation analysis, copy the registration data from the reg directory of the feet analysis into the correlation run. Run a higher level analysis by combining runs within each subject. First, select higher level analysis and then stats plus post stats.

Then within the data tab, choose inputs are lower level feet directories and enter the subject's runs within the stats tab. Choose mixed effects. Simple OLS set up a model as the mean effect and enter a value of one for each of the subjects runs.

To combine data overrun between subjects an ordinary least square simple mixed effects analysis should be used for this, choose higher level analysis and stats plus post stats within the data tab. Choose inputs are lower level feet directories and enter the subject's combined Runs within the stats tab, choose mixed effects. Simple OLS set up a model as three groups enter a value of one for the group.

Each subject belongs to zero. Otherwise, group analysis should be done on each voxel using a one-way innova with three levels which correspond to the three groups to threshold. The Z statistic images use a cluster forming threshold of Z greater than 2.0 and a corrected cluster significant threshold of P equals 0.05 to obtain correct Z values on the correlation map.

A reverse fisher Z transform should be performed on the results. Finally, use the following specific contrasts as seen on the screen here. This figure shows the default mode network revealed with connectivity from a posterior seed, including the retro splenium and precuneus in red yellow colors and in anterior seed, including the vent medial prefrontal cortex in blue green colors.

The first row reveals the network for control subjects, the second row for left temporal lobe epilepsy, and the bottom row for right temporal lobe epilepsy. The following figures compare these networks between these three groups. Here we see the default mode networks revealed with an anterior and a posterior seed for combined right and left temporal lobe epilepsy compared to healthy controls.

This figure shows the default mode networks revealed with the same seed points for left temporal lobe epilepsy only compared to healthy controls. While this figure shows the networks revealed for right temporal lobe epilepsy only compared to healthy controls, and finally here we see the default mode networks revealed with an anterior and a posterior seed for left temporal lobe epilepsy compared to right temporal lobe epilepsy. Studies of functional connectivity that include the entire brain are essential for understanding the fundamental mechanisms of epilepsy.

We used a seed based technique in this experiment to assess connectivity to the default mode network. It'll be interesting to see how other techniques compare in their results when studying temporal lobe epilepsy.