SEEG shaft-tip offsets share a direction within implants beyond a per-shaft rotation null

We tested whether shaft tips in one stereoelectroencephalography (SEEG) implant deviate from bolt axes in a common direction. The analysis used OpenNeuro ds005083 v1.0.0 bolt and sorted-contact files from 60 implants in 58 pediatric patients, with 611 shafts in native post-operative CT space. A plan was written and hashed before deviation vectors were computed. For each shaft, we measured the deepest contact's direction away from the extended bolt axis. For each implant, R was the length of the mean unit direction. The null rotated each direction independently around its own bolt axis, preserving each shaft's axis and offset size. Across implants, mean observed R was 0.386 versus a null mean of 0.290 (one-sided group p = 0.0002, the smallest possible value with 5000 draws). Forty of 60 implants had R above their null mean, and 13 had per-implant p < 0.05. Results remained significant for shafts with offsets >= 1 mm (47 implants; p = 0.0002), for a post hoc null that retained the pooled direction (p = 0.0002), and when one implant per patient was used (58 implants; p = 0.0002). The pooled direction was mostly superior and partly anterior in native CT axes; planned tests found no posterior or inferior direction (p = 0.98 and 1.0). These findings show common within-implant directions beyond independent bends and a pooled superior component, but do not identify the cause. CT axes are not in standard space, the .dat x sign is uncertain, public surgical plans, prior-craniotomy labels, and CT timing are absent, and manual bolt-point placement could cause a common bias. The analysis therefore cannot establish whether brain shift, insertion force, or localization error caused the directions.

stereoelectroencephalography · SEEG · electrode trajectory · bolt axis · brain shift · CT localization · directional statistics · Monte Carlo test