Step 3: Calculation of Coupling Measures

In this step, we will be calculating coupling measures between electrodes to determine any temporal correlations.

This allows us to create an adjacency matrix representation of a network graph that we can calcalate network summary statistics on. For more information regarding network graph representations, please check Network Graph Representations.

We support three different correlation measures: the linear Pearson’s Correlation, and non-linear mean phase coherence correlation, and the non-linear correlation coefficient (Pijn’s Measure).

When you get to the Correlation page, you will see the following form. This page works very similarly to the conversion page.

_images/correlation_form.png
Correlation Overall Input Parameters
Input Description
Patient ID The patient ID from the patient demographics page
Path to CSF Files The full path to the CSF fules produced in step 1.
Coupling Measures The list of coupling measures you’d like to select

Once you have selected the location of the CSF files (for one patient), you can select as many event windows as you’d like by clicking Add Seizure Event and Remove.

Correlation Event Window Input Parameters
Input Description
Event Start Time The start time of the event
Event Start Time The end time of the event
Lag Window The lag window to calculate over
List of Channels The list of electrodes over which you’d like to calculate correlations
Path to Correlation Metrics Output File The output file path for each event (must be unique for each entry)

Once you click Run Correlations you will be taken to a very similar screen that allows you to view your status as shown below:

_images/correlation_thanks.png

Again, a popup will take you to the next step, network analysis.