Skip to content. | Skip to navigation

Sections
Personal tools
You are here: Home Forum Using data sets HFO marking label for ieeg-1 (intracranial electroencephalography 1)

HFO marking label for ieeg-1 (intracranial electroencephalography 1)

Up to Using data sets

HFO marking label for ieeg-1 (intracranial electroencephalography 1)

Posted by Mostafa Mohammadpour at September 30. 2019

Hello everybody,


I am working on HFO detection algorithm on iEEG data and I used "ieeg-1 (intracranial electroencephalography 1)" dataset.
In order to validate algorithm we need to have the dataset ground truth (HFO marker).
There are three different HFO marked label that come from their proposed algorithm, but I need real HFO marker which come from reviewers.
 
Best regards,
Mostafa

Re: HFO marking label for ieeg-1 (intracranial electroencephalography 1)

Posted by Johannes Sarnthein at October 11. 2019

Hi Mostafa,

The prospective definition of a clinically relevant HFO is based on the automated detection. (We only review events visually to reject artifacts.)

Simultaneous fast ripples and ripples (FRandR) had the highest clinical relevance as defined by their power to predict seizure freedom.

So FRandR are the ground truth. 

You should train your algorithm on FRandR. Make sure to benchmark your results against seizure freedom.

Best,

Johannes

 

 

Re: HFO marking label for ieeg-1 (intracranial electroencephalography 1)

Posted by Mostafa Mohammadpour at October 16. 2019

Hi Johannes,

Are there any meta-data available for this dataset? like SOZ, Resected area, or seizure freedom data?

 

Best,

Mostafa

Re: HFO marking label for ieeg-1 (intracranial electroencephalography 1)

Posted by Johannes Sarnthein at October 17. 2019

Hi Mostafa,

Please have a look at our publication https://www.nature.com/articles/s41598-017-13064-1

Table 1 lists the resected area (RA) anatomically and the postsurgical seizure freedom (ILAE scale) for each patient.

In more detail, the supplementary material (Table 1S) lists all channel locations that were resected (highlighted in red). https://doi.org/10.1038/s41598-017-13064-1

The seizure onset zone (SOZ) as defined by the epileptologists is included in the resected channels.

For easy access, I attach the supplementary Table 1S here.

All the best,

Johannes

 

Fedele T, Burnos S, Boran E, Krayenbühl N, Hilfiker P, Grunwald T, Sarnthein J. Resection of high frequency oscillations predicts seizure outcome in the individual patient. Nature Scientific Reports. 2017;7:13836.

 

 

Attachments
Powered by Ploneboard
Document Actions