Hello Brainstorm team,
I’d like to average Epileptogenicity maps across subjects, but SEEG coverage differs between them. What is the recommended way to control for spatial sampling bias—e.g., restricting analysis to sampled parcels only, using coverage-aware weighting, or doing ROI-level aggregation plus a meta-analytic combine?
Any suggested workflows or example scripts in Brainstorm would be greatly appreciated.
Thanks in advance!
Best regards
Hi @Renoald,
You can perform analysis of sEEG data across multiple subjects with the MIA toolbox, which is available as standalone and as a Brainstorm plugin.
MIA – A.-Sophie Dubarry
Redirecting
Indeed, MIA's developer is in the forum. Ping, @as_dub
Thank you very much for the suggestion!.
I have a question for the MIA engineers @as_dub .I have installed the MIA toolbox. However, I’m facing a new issue: I already have the Epileptogenicity Map for each subject. Is there a way to directly call MIA functions or scripts to compute a cross-subject (group-level) average?
Best regards
Hi @Renoald MIA allows exploring time series of several patients within region of interests. Not sure if it would work with Epiletogenicity Map. What type of object/data are those in Brainstorm? Is it time resolved?
Best,
AnneSo
Thank you very much for your reply.
In Brainstorm, I have completed all preprocessing for each patient and each seizure, including time–frequency analysis, baseline normalization, and the computation of an Epileptogenicity Map (EM) per seizure. The EM outputs I have are time-resolved around seizure onset. As shown in the attached screenshot, I can obtain:
ROI (scout) time series: baseline-normalized EM values per ROI across time (per seizure, per subject).
Source/volume maps: EM as dynamic source maps that can be exported as 4D NIfTI (or as a sequence of time-stamped volumes
The data structure is as follows:
I was wondering whether I could call MIA’s scripts or functions to perform a group-level analysis on my data. Many of the steps are repetitive; if you could let me know which parameters I should pass to the relevant functions, I would be very grateful.
Best regards,
Renoald
Besides, my primary concern is whether MIA can account for (or mitigate) the issue of uneven sEEG spatial coverage across subjects.This has been a long-standing concern for me.
Hi @as_dub ,
I imported two patients from Brainstorm into MIA and created a study including both. However, in Display → Show table I can still operate on only one patient. The study .mat confirms both patients are present (screenshot attached).
How do I enable multi-subject interaction in the table?
Also, what is the recommended way to average EM across subjects, especially with uneven sEEG coverage ?
Thanks in advance!
Best regards
