SEEG Time-Frequency Fingerprint Analysis for Epileptogenic Zone Localization

Dear Brainstorm Software Engineering Team,

I am writing to seek your technical support regarding an issue I encountered while following the official workflow for SEEG Time-Frequency Fingerprint Analysis for Epileptogenic Zone Localization. I have strictly executed every step of the standard pipeline as documented, yet I cannot obtain the expected output at the Compute inverse stage.

Per the official tutorial page: https://neuroimage.usc.edu/brainstorm/seeg/SeizureFingerprinting#Modeling_interictal_spikes, I configured all inverse calculation parameters exactly as illustrated in the guide and ran the computation. However, the pipeline fails to generate the required inversion kernel and LINK files; the only output produced is a volumetric grey brain mesh file instead of the cortical surface LINK data needed for downstream fingerprinting.

I have identified a critical discrepancy in functionality: running the inverse calculation directly on the raw signal file (Link to raw file) successfully yields both the inversion kernel and valid LINK files without any errors. In contrast, when processing segmented epoch data imported into the Brainstorm database, the workflow reverts to volumetric source reconstruction exclusively.

Additionally, I am unable to locate the epilepsy fingerprint toolset under the Process Selection menu, and I lack clear instructions on how to properly load or activate the Seizure Fingerprinting plugin within the software environment.

This barrier halts my complete interictal spike fingerprint analysis pipeline, and I would greatly appreciate your prompt troubleshooting guidance to resolve this limitation. Thank you for your time and assistance.

Best regards,

Alice Wang

Hello Alice,

Did you get any message either in the GUI or in the Matlab command window?
If only one file is returned, it could be the case that the option Output mode was set to Full results (Kernel*Recordings)

The sources are computed in the source space of the head-model. Double-check that the head model source space is Cortex surface:
https://neuroimage.usc.edu/brainstorm/seeg/SeizureFingerprinting#Head_modeling

Not sure which epilepsy fingerprint toolset under the Process Selection menu your refer to.
Can you clarify this?

There is no Seizure Fingerprinting plugin in Brainstorm, in the context of epileptic seizures, fingerprinting refers a distinct time-frequency pattern (named “fingerprint”) that characterizes the epileptogenic zone.

Please note that the SeizureFingerprinting tutorial is the third and last part of a series of tutorials, so you need to complete the other two tutorials before hand:

  1. CT to MRI co-registration
  2. Contact localization and labeling
  3. Time-Frequency Fingerprint Analysis for Epileptogenic Zone Localization