Theta closed-loop stimulation did not increase distant SEEG evoked potentials more than control in five analyzable sessions

We measured single-pulse evoked potentials (EPs) before and after theta closed-loop (CL) and control stimulation blocks in OpenNeuro ds006065 v1.0.0 (9 sessions, 7 subjects). The files have no event markers. Thus, we found the pulse times and the stimulated contacts from the stimulation artifact. We used bipolar pairs more than 15 mm from the stimulated contacts and selected responsive pairs symmetrically on pooled pre and post trials. The outcome was the median log2(post/pre) EP RMS amplitude, with a pulse-label permutation null. Only 5 sessions had 3 or more responsive distant pairs. The CL changes (-0.83, +0.23, -0.27 log2) were not larger than the control changes (+0.63, +0.32). Exact Mann-Whitney p = 0.2. In both within-subject pairs (p17, p19), the CL change was smaller than the control change. Only p18 CL had a significant increase (+0.23 log2, p = 0.016). A 5 mm exclusion radius gave the same conclusion (p = 0.4). The pre vs post rank correlation of the EP amplitude map (median about 0.78) was near the split-half reliability within the pre run (about 0.84). Thus, one stimulation block caused map changes that were small relative to the measurement noise. Limits: the sample is small, the ISI is not the same in all sessions, and few distant pairs respond. The package contains code, hashes, tables and figures.

single-pulse evoked potentials · stereo-EEG · closed-loop stimulation · hippocampal stimulation · evoked-potential maps · OpenNeuro · permutation test