Evaluating dipolar source localization feasibility from intracerebral SEEG recordings

Area of Science:

• Neuroscience
• Medical Imaging
• Epileptology

Background:

• Stereo-electroencephalography (SEEG) is crucial for pre-surgical evaluation in drug-resistant epilepsy.
• Current SEEG analysis uses bipolar montages, limiting the view of distant neural activity propagation.
• Electrical source localization (ESL) methods are primarily developed for surface EEG/MEG, with limited exploration on SEEG data.

Purpose of the Study:

• To investigate the feasibility and accuracy of Electrical Source Localization (ESL) using common reference SEEG signals.
• To explore the potential of ESL for identifying neural generators and distant brain activity propagation.
• To evaluate the influence of volume conduction models, sensor configuration, and noise on ESL accuracy in SEEG.

Main Methods:

• Utilized common reference SEEG signals, preserving volume propagation information.
• Developed and tested ESL approaches, including equivalent current dipole models and analytical volume conduction models.
• Validated methods using realistic simulations, real SEEG data from intracerebral electrical stimulations (known sources), and epileptic interictal spikes.

Main Results:

• Common reference SEEG signals allow for ESL, capturing volume propagation information.
• A straightforward ESL approach using equivalent current dipole models achieved localization accuracy below 10mm under specific conditions.
• The study analyzed the impact of volume conduction models, SEEG sensor spatial configuration, and noise levels on localization precision.

Conclusions:

• Electrical source imaging using common reference SEEG signals is a promising technique for investigating neural generators.
• This method can accurately localize sources, offering a valuable complement to routine visual interpretation in epilepsy surgery planning.
• The findings support the use of ESL with SEEG for enhanced pre-surgical evaluation in drug-resistant partial epilepsy.