Validating Patient-Specific Finite Element Models of Direct Electrocortical Stimulation

Area of Science:

• Neuroscience
• Biomedical Engineering
• Computational Modeling

Background:

• Direct electrocortical stimulation (DECS) is used for epilepsy and emerging applications like stroke rehabilitation and brain-computer interfaces.
• Understanding the electrophysiological mechanisms of DECS requires accurate computational models.
• Accurate electrode localization in patient-specific models is challenging due to brain shift post-implantation.

Purpose of the Study:

• To validate patient-specific computational models of DECS against in vivo voltage recordings.
• To quantify the impact of electrode localization parameters on simulated cortical surface voltages.
• To identify optimal electrode localization parameters for DECS modeling.

Main Methods:

• Intracranial voltages were recorded during DECS in six epilepsy patients.
• Computational models incorporated varying electrode projection methods (principal axis, Hermes, Dykstra) and cerebrospinal fluid (CSF) depths (0, 1, 2 mm).
• Model accuracy was assessed by comparing simulated voltages to in vivo recordings.

Main Results:

• Increased CSF depth between the electrode and cortical surface led to higher model errors and reduced voltage prediction accuracy.
• The Hermes projection method with 0 mm CSF depth provided the best estimation of recorded voltages (r=0.92, slope=1.21).
• These findings quantify the influence of electrode localization parameters on DECS model accuracy in vivo.

Conclusions:

• Electrode localization parameters significantly impact the accuracy of computational models for DECS.
• The Hermes method with 0 mm CSF depth is recommended for patient-specific DECS modeling.
• This validation provides a foundation for future research into DECS mechanisms and applications.