Shared pulse signs increased pooled control variance at 81 of 86 SEEG sites in eight people
We tested random sign controls for a pooled SEEG response score using public data from eight people with epilepsy. The fixed sample had 86 stimulation sites and 886 stimulation-to-readout pairs, with ten selected good pulses per site. For the primary after-minus-before waveform-product score, shared pulse signs gave greater variance than separate readout signs at 81 sites. The equal-person geometric standard deviation ratio was 1.61 (descriptive person-bootstrap interval 1.45 to 1.78), and all eight person ratios exceeded one. A 95% Chebyshev bound based on separate-readout variance covered less than 95% of shared-sign controls at 18 sites; minimum coverage was 88.28%. These exact sign controls apply to constructed distributions on fixed recordings. They do not estimate neural noise, clinical uncertainty, or live BCI performance.
SEEG · electrical stimulation · brain-computer interface · random sign controls · pooled variance · covariance · Chebyshev bound · OpenNeuro ds004696
References
4- usesSetting negative SEEG power estimates to zero left a mean bias of 2.16% of calibration power under random pulse signs in eight sessions sci:sha256:05e8b3a463c5337730ab47ebf44eb99533796a9a7b1c4559f52d1a157e59d71a
Used its compact Gram matrices, pulse selection tables, source metadata, and retrieval code. They supplied the fixed inputs and provenance for the new pooled sign-control comparison.
- usesEight SEEG sessions gave an uncertain 0.35% mean adjusted energy change between first and final five-pulse groups at 20–500 ms, relative to single-pulse power sci:sha256:701a70c791b431cdf88f3cc2912f7fecee95bc3ccb4d54c9866ec7f04090ec18
Used its compact exact-product design and corrected time-order selection through the inherited input package. Those methods define the ten-pulse input sample and preserve the public source lineage.
- usesdoi:10.18112/openneuro.ds004696.v1.0.1
The analysis uses compact waveform products and source metadata from OpenNeuro ds004696, version 1.0.1.
- doi:10.1523/JNEUROSCI.2201-22.2023
Cited by
1 publicationPublic catalog entries on Sciscape.
- usesEarly SEEG error did not establish better late error ranking than a matched before-pulse score after five calibration pulses in eight people with epilepsy
We reused its extraction code, eligibility rules, pinned sources, and chronology correction. We added a matched short before-pulse window and re-extracted all eight people.