Knowledge-based automated planning system for StereoElectroEncephaloGraphy: A center-based scenario.

Area of Science:

• Neurosurgery
• Medical Imaging
• Computational Anatomy

Background:

• Stereo-ElectroEncephalography (SEEG) surgical planning is complex and varies between institutions.
• Existing methods rely heavily on individual surgeon experience and institutional workflows.
• A standardized, data-driven approach is needed to improve planning consistency.

Purpose of the Study:

• To develop and validate a data-based Computer Assisted Planning (CAP) system for SEEG.
• To leverage retrospective data for generating and optimizing patient-specific surgical trajectories.
• To improve the efficiency and consistency of SEEG surgical planning.

Main Methods:

• Developed a CAP solution utilizing retrospective patient data to propose common trajectories.
• Implemented an optimization framework to adapt trajectories based on patient anatomy and clinical requirements (e.g., distance from vessels, gray matter recording, insertion angle).
• Validated the system with 15 retrospective SEEG cases, reviewing over 200 trajectories with two neurosurgeons.

Main Results:

• The CAP system proposed institution-specific trajectories, adapted through an optimization framework.
• Neurosurgeons deemed approximately 81% of optimized trajectories clinically feasible, with 75% inter-rater reliability.
• Optimized trajectories demonstrated superior or comparable quantitative metrics (distance from vessels, insertion angle, gray matter recording) compared to manual plans.

Conclusions:

• A tailored, data-based CAP solution can effectively generate clinically feasible SEEG trajectories.
• The system shows potential to increase the acceptance rate of automated trajectories in SEEG planning.
• This approach offers a standardized, data-driven method to assist neurosurgeons in SEEG planning.