Towards a mesoscale physical modeling framework for stereotactic-EEG recordings.

Area of Science:

• Computational neuroscience
• Biophysics
• Neuroimaging

Background:

• Mesoscale neural mass population models (NMMs) are used to model stereotactic-electroencephalography (SEEG) and scalp EEG recordings.
• The physical basis connecting NMMs to measurement physics, especially at the intermediate spatial scale of SEEG, remains unclear.
• Combining NMMs with volume conductor models for SEEG data representation is challenging.

Purpose of the Study:

• To develop a framework that integrates multi-compartmental neuron models with NMMs and volume conductor models.
• To simulate transmembrane currents and current source density (CSD) profiles within cortical layers.
• To accurately model SEEG signals by bridging microscale neuronal activity and macroscopic electrophysiological recordings.

Main Methods:

• Utilized a multi-compartmental modeling formalism combined with a detailed geometrical model.
• Simulated transmembrane currents in pyramidal cells (layers 3, 5, 6) due to synaptic input.
• Developed a framework linking NMM activity to a volume conductor model to simulate SEEG signals.

Main Results:

• Current source density (CSD) is highly sensitive to the distribution of synaptic inputs across cortical layers.
• Equivalent current dipole strengths vary significantly based on input location, impacting electrophysiological recordings.
• Raw NMM outputs are not direct proxies for electrical recordings; a simplified CSD model with layer-specific sources can accurately reproduce SEEG measurements.

Conclusions:

• Accurate modeling of electrophysiological recordings requires incorporating physical measurement models into NMMs.
• The developed framework connects microscale neuron models with NMMs and physical measurement models.
• This approach enhances the accuracy of predicted electrophysiological recordings, improving the interpretation of SEEG and EEG data.