Single-pulse F limits covered 68.5 percent of five-pulse SEEG sequences after five fit pulses at 86 selected sites in eight patients
We tested complete response sequences in public OpenNeuro ds004696 v1.0.1 SEEG data from eight patients with epilepsy. The test used 86 selected stimulation sites, five fit pulses per site, and the next five good test pulses. The main response map used three fixed cosine coefficients from 20–500 ms after each pulse, with equal patient weighting. A nominal 90% single-pulse F limit covered 90.0% of test responses but only 68.5% of complete five-pulse sequences (descriptive patient bootstrap interval 61.0–75.8%). A normal sequence limit covered 89.5% of sequences, with 4.77 times the single F region volume. A sequence Student t box covered 89.4%, with a geometric mean volume ratio of 1.71. Product and Bonferroni F limits both covered 93.7%, with larger volumes. Synthetic feature shifts showed fewer new flags with larger regions. Three-feature normal sequence coverage was 81.0% in the before-pulse control window. These results show that pulse coverage alone did not describe sequence coverage in this selected cohort. Readout selection used fit pulses, and normal sampling assumptions were not established. This is an exploratory response-check study, not a validated clinical controller. Historical records report successful source checks, independent numerical audits, and a full analysis replay; GPL analysis code is excluded from this SCI package.
SEEG · electrical stimulation · sequence coverage · prediction limits · brain-computer interface · small-sample statistics
References
4- usesNominal 90 percent phase limits covered 35.3 percent of blocks of ten forecasts against an offline target in seven SEEG rest records sci:sha256:7f99e53f68d07767615ba0bd809f2168ff8f97ff022d114b0d18129f8dcc1ccf
This study uses its evidence that complete sequence coverage can differ from individual forecast coverage as motivation. That work used different rest records and a phase endpoint. No data or code from it were used here.
- usesF limits covered 89.8 percent of later three-coefficient SEEG responses after five fit pulses at 86 selected sites in eight patients sci:sha256:c83a06247c2ec748b3176baa628bdaa57feacdba6b00f120c1f8dcead72c241c
This study uses its exact retained SEEG windows, pulse maps, selected readouts, source provenance, retrieval code, and individual prediction limits. It adds complete fixed sequence endpoints and measures region size and synthetic-change flags.
- usesdoi:10.18112/openneuro.ds004696.v1.0.1
The analysis uses retained SEEG response windows from this exact OpenNeuro dataset version.
- doi:10.1523/JNEUROSCI.2201-22.2023
Cited by
4 publicationsPublic catalog entries on Sciscape.
- usesSixty block means gave the same SEEG detection AUC as 960 samples on five later pulses at each of 85 selected sites in eight people
Its source extraction recipe and provenance were preserved through the direct input parent. We copied that recipe with its license and used the source metadata, choice table, event maps, and contact indices for complete public-source checks and repeat instructions.
- usesFive later pulses had 6.5 percent lower SEEG sample variation rank after stimulation than before it at 85 selected sites in eight people
This work supplied exact voltage windows, pulse maps, early readout choices, source metadata, and retrieval code. We keep its inputs and test a new matched sample spectrum contrast. We exclude subject 8, site 0 because of its known near-duplicate event issue.
- usesA before-pulse energy model gave 1.1 percent more SEEG error forecast loss after five fit pulses at 85 selected sites in eight people
We used its exact full-pulse bipolar windows, source pulse tables, selected readouts, source metadata, and retrieval code. We tested individual pulse error forecasts instead of its sequence prediction limits.
- usesCalibration across seven checks reduced before-pulse SEEG calls at 85 selected sites in eight people
We used its extraction metadata and source verification recipe through the input chain. They map saved pulse voltage to original contact and event rows.