Compute an epileptogenicity map
How to compute maps of epileptogenicity from ictal iEEG recordings in Brainstorm - showing where in the brain the high-frequency activity that characterizes seizure onset is concentrated.
Starting from seizure recordings and known electrode positions, you compute the Epileptogenicity Index and project it onto the patient's anatomy, producing a spatial map of epileptogenicity.
The methods and software used in this tutorial have not been certified for clinical practice, and should be considered for research purposes only.
What the Epileptogenicity Index measures
The Epileptogenicity Index quantifies the shift towards high-frequency activity that marks the onset of a seizure, and how early it appears at each recording site. Computed across all contacts and projected onto the anatomy, it gives a spatial picture of which regions are driving seizure onset rather than following it.
What you will need
Applications - Brainstorm, from the Applications catalogue.
Data - the Epimap tutorial dataset, which contains everything this tutorial uses: pre- and post-implantation T1 MRI scans, the electrode positions, and three seizure recordings (SZ1, SZ2, SZ3). Copy it into your own workspace first.
Prior work - electrode positions have to be established before you start. Place SEEG electrodes with Brainstorm covers that, and the Epimap dataset also ships with positions already prepared.
The workflow
Set up the subject
Import the anatomy and the electrode positions, so recordings can be related to locations in the brain.
Import the seizure recordings
Load the three ictal recordings and mark the onset of each seizure.
Define baseline and seizure windows
Set the pre-seizure baseline and the seizure period the index compares against each other.
Compute the index
Run the epileptogenicity computation across all contacts for each seizure.
Map it onto the anatomy
Project the per-contact values onto the patient's own anatomy to see the spatial distribution.
Compare across seizures
Look at the three recordings together - agreement across seizures is what makes a result convincing.
Where to go next
Compute CCEP responses works with stimulation rather than ictal data, on the same platform tooling.