Pulse sign tests across six contacts gave 9% call rates before stimulation in two SEEG recordings
We tested whether a response rule can account for a search over time and six SEEG contacts. We used 630 recorded stimulation pulses from OpenNeuro ds004774 version 1.0.0. The data came from 63 fixed surface stimulation sites in two recordings. Each site had ten pulses and six depth readout contacts. The primary test changed whole pulse signs in all 1,024 patterns. It used the maximum across time and contacts in a 20–90 ms window. Across four matched control windows, call rates fell from 92.4% and 82.5% for point t tests to 9.1% and 9.2%. A time-only sign test had rates of 28.8% and 26.7%. The primary test detected 87.0–93.0% of expert response labels in three expert-record groups. Specificity ranged from 50.7% to 76.4%. Bonferroni gave fewer control calls but lower sensitivity. A finite-sample proof gives a 5% site error bound only under joint pulse sign symmetry. The observed rates do not establish that bound for the source noise. The package preserves input arrays, labels, pinned source versions, range hashes, exact p-values, code, figures, and repeat steps. A fresh rebuild of all 630 excerpts and a separate numeric replay passed. These are offline results from two people.
SEEG · evoked responses · sign-flip test · multiple testing · negative controls · expert labels · brain-computer interfaces
References
4- usesTen-pulse means had higher response label AUCs than first pulses for a fixed 10–90 ms absolute score in two SEEG recordings sci:sha256:1dfa7a52fe9acf5f4a7a85e2004bf3a19d22575d694466d89d391e4150e0d226
We used the fixed two-recording SEEG pulse arrays, source labels, metadata, event selection, range hashes, and retrieval code. These inputs enabled the exact time and contact search tests without a new data selection.
- usesdoi:10.18112/openneuro.ds004774.v1.0.0
The analyzed arrays, expert labels, source metadata, and selected events are from OpenNeuro ds004774 v1.0.0; the fixed excerpts and source provenance are preserved in this package.
- doi:10.1016/j.jneumeth.2025.110389
- doi:10.1016/j.neuroimage.2014.01.060
Cited by
2 publicationsPublic catalog entries on Sciscape.
- usesA 200 ms baseline gave 53 to 61 percent control calls in two SEEG recordings at a nominal 5 percent limit
We used its description as prior work on null tests across six SEEG contacts. Our test uses baseline segments and different windows and score correction. We do not infer a method advantage from rates across these different targets.
- usesRemoval of shared voltage reduced site identification accuracy for a fixed method in two SEEG recordings with six contacts
This publication supplied the unchanged SEEG epochs, trial timing, source metadata, pinned range hashes, and retrieval code. We checked the used raw ranges again and applied a new site identification analysis.