Seizure onset analyses using the Epileptogenicity Index
Comment from Pr Jonathan Curot,
Neurologist at Toulouse University Hospital, France.
“Precise localization of the epileptic zone (EZ) from intracranial EEG (iEEG) recordings is of utmost importance, and remains a major challenge in epilepsy surgery.
In recent years, researchers have developed several methods to improve EZ localization, one of which is the epileptogenicity index (EI). The EI is a powerful analytical tool that transforms qualitative, expert-driven interpretation into a quantitative framework for identifying epileptogenic brain regions.
The EI computes an energy ratio (high vs. low frequencies) and detects change points at seizure onset to assign a continuous score to each channel, ranking them based on both the intensity and the earliness of their recruitment (Bartolomei et al., Brain, 2008).
As a clinician, I see clear value in the use of the EI in signal analysis, as it provides a standardized approach with reproducible metrics that can meaningfully serve decision-making. Its standardization is however limited by its dependence on preprocessing choices (frequency bands, thresholdings, etc), and, above all, by its reliance on the accurate identification of seizure onset time. In addition, it does not account for connectivity between regions, which represents another limitation.
Nevertheless, when combined with other biomarkers, the EI appears to improve the precision of EZ identification, and is associated with good prediction of surgical outcomes and seizure freedom (Makhalova et al., Ann Clin Transl Neurol., 2023). It also serves as a useful adjunct for identifying potential thermocoagulation targets.
To date, the EI has been used by only a limited number of expert groups. Its integration into SEEG analyses software could increase accessibility for research groups, enabling broader evaluation of its relevance within routine analysis workflows."
Pr Curot is an epileptologist practicing at the Toulouse University Hospital, and working in research at the Brain and Cognition Research Center (CerCo) CNRS lab (Toulouse, France). He has a long-standing experience in intracranial EEG quantitative analysis, from both macro and micro electrodes recordings.