Deep SEEG contacts deviate from the anchor-bolt axis in a pediatric cohort

We measured the perpendicular distance of stereoelectroencephalography (SEEG) contact centres from the extended anchor-bolt axis in public post-implant CT trajectory derivatives for 58 pediatric patients (60 subject-sessions, 613 shafts). The deepest contact was a median 1.15 mm from the bolt axis (95% subject-bootstrap CI 1.03–1.24 mm); 21% of shaft tips were more than 2 mm away and 9.5% more than 4 mm away. Deviation increased with contact depth in all 58 subjects (mean subject median Spearman rho 0.80; one-sided sign-flip p=1e-4, the smallest possible value with 10,000 flips). This result is consistent with a shaft angle and does not establish bending. Median shaft-to-bolt angle was 1.63 degrees and median curvature was 0.37 mm. A post hoc exclusion of 18 shafts with angle >15 degrees or curvature >2 mm yielded median tip deflection 1.12 mm and 7.1% of tips above 4 mm. These data indicate that the bolt axis is not a reliable proxy for deep contact location in this cohort; post-implant imaging is needed. The public files contain no planned trajectories, so the analysis measures deviation from the bolt axis, not targeting error against a plan. The cohort is pediatric and from one center; segmentation errors may contribute to extreme values.

stereoelectroencephalography · SEEG · electrode localization · anchor bolt · trajectory · pediatric · computed tomography · brain-computer interface