Patient-Specific Sensor Registration for Electrical Source Imaging Using a Deformable Head Model.

Area of Science:

• Neuroscience
• Biomedical Engineering
• Medical Imaging

Background:

• Accurate electrical source imaging (ESI) relies on individualized bioelectric head models.
• Constructing these models necessitates precise scalp electrode placement.
• Current methods like photogrammetry are user-intensive and hinder clinical integration.

Purpose of the Study:

• To introduce and validate a novel, fully-automatic method for sensor registration.
• To overcome limitations of manual methods in clinical settings.
• To enhance the efficiency and reproducibility of ESI.

Main Methods:

• Utilizing a shape-constrained deformable head model for scalp segmentation.
• Registering average electrode coordinates to a mean scalp mesh.
• Employing point-based correspondence on the adapted scalp surface for patient-specific localization.

Main Results:

• The proposed automated method achieves accuracy comparable to state-of-the-art photogrammetry.
• Quantified electrode variability demonstrates the robustness of the new approach.
• Simulated and real data validated the performance of the sensor registration technique.

Conclusions:

• A fully-automated, model-based sensor registration is feasible and accurate.
• The method eliminates user-dependent inaccuracies, ensuring reproducible results.
• This automation facilitates clinical workflow integration, promoting wider ESI adoption.