Zestaw narzędzi electro-MICA do integracji elektrofizjologii w multimodalnych obrazowaniach i analizach łączeń mózgu
The electro-MICA toolbox for integrating electrophysiology within multimodal imaging and connectomics workflows
W skrócie
[Preprint - wstępne wyniki] Naukowcy stworzyli bezpłatny program komputerowy electro-MICA, który łączy pomiary elektrycznej aktywności mózgu (zarówno ze skóry głowy, jak i z głębokich elektrod) z obrazami mózgu z innych badań. Program wykorzystuje zaawansowane metody matematyczne do precyzyjnego określenia, gdzie w mózgu pochodzi zarejestrowana aktywność elektryczna. Testowanie na danych pacjentów z epilepsją pokazało, że ta nowa metoda jest bardziej dokładna niż dotychczasowe podejścia, a program nie wymaga od użytkownika żadnych skomplikowanych ustawień.
Oryginalny abstract (angielski)
The integration of electrophysiological recordings with multimodal neuroimaging data holds great promise for advancing our understanding of brain function and neurological disorders. To facilitate this endeavour, we present electro-MICA, an open-access Python toolbox designed to project electrophysiological features from scalp and intracranial electroencephalography (EEG) onto cortical and hippocampal surfaces generated by validated multimodal imaging ecosystems. The toolbox comprises two pipelines: one for intracranial EEG (iEEG) recorded with stereo-EEG depth electrodes, and one for scalp EEG source localization. Both pipelines are grounded in numerical solutions to the electromagnetic equations governing electric activity in the brain, solved using the Boundary Element Method. A key methodological contribution is the use of a current density double layer model for neural generators, which avoids the mathematical singularities introduced by conventional dipole-based models when electrodes are near the cortical surface, a situation that can arise in iEEG. Electrode contacts are additionally modeled with non-zero length, improving physical realism. Scalp EEG source localization is performed using eLORETA on a subject-specific three-layer head model derived from the anatomical input. Validation against empirical gamma-band iEEG data from 32 subjects demonstrates that the distributed generator model outperforms both distance-based and dipole-based alternatives. An illustrative clinical example demonstrates the toolbox's capacity to reveal associations between intracranial spike rates, cortical thickness, and anatomical connectivity in an epilepsy patient. Electro-MICA requires no parameter selection from the user, facilitating straightforward multimodal analyses in both research and clinical settings. The toolbox is available at github.com/MICA-MNI/electromica with extensive online documentation at electromica.readthedocs.io.