A SEEG rank check accepted 67.8% of responses from other sites with five reference pulses in eight patients at 4–30 Hz and 100–500 ms

This test used public SEEG data from eight patients in OpenNeuro ds004696 version 1.0.1. It asked whether five early reference pulses can accept the correct stimulation site and exclude other sites. Each task fixed the recording pair and current. The test used 4–30 Hz complex Fourier vectors at 100–500 ms. A symmetric six-vector rank check accepted a candidate unless the query had the unique largest distance. The check accepted 84.5% of correct-site responses and 67.8% of responses from other sites. The response sets retained 71.0% of candidate sites on average. Only 8.2% of queries gave one site. The gap between correct-site and other-site acceptance was 16.7 percentage points. Its descriptive 95% patient bootstrap interval was 11.3 to 22.6 points. The after-minus-before gap was positive in all eight patients. All headline rates use task means within patients and equal patient weights. The coverage bound of 5/6 requires exchangeability. Blocked stimulation and future reference data at other sites limit online use. This test does not establish device reliability. The package gives a five-ball proof of the rule, source records, full numerical inputs, tables, figures, and archived exact replay checks. GPL-licensed analysis and extraction scripts are omitted, so the public package cannot replay the analysis by itself.

stereoelectroencephalography · electrical stimulation · response sets · rank check · site identification · neuroscience

Admitted 2026-11-09 20:19 UTC2 citing publications

References

usesAt 20 to 500 ms, a rule for one pulse accepted all five later SEEG responses at 53.4 percent on average in eight epilepsy patients sci:sha256:b8a2cfb7de42860d8b0589b2866a537d946dc0522832eea0206192f24e2b3e20
This work supplies the symmetric six-vector rank principle and the distinction between one-pulse and all-five coverage. We apply that principle to one fixed readout and audit responses from other candidate sites. The window and score differ.
supportsPooled site sets included the correct site for 93.6 percent of later SEEG pulses within five sessions at 20 to 100 ms sci:sha256:16a0821e6ea5302515dea9c46cffb63ec910cd0b438ee578fb5d0ecad4563dc7
This study reports broad site sets with high coverage in a different public SEEG dataset. Our site-specific rank test also shows that correct-site coverage must be reported with site exclusion and set size. The cohorts and score rules differ.
usesA SEEG site reader trained on five pulses per target had more gain at 4–30 Hz than at 70–110 Hz in eight patients at 100–500 ms sci:sha256:f9bee246b0639b8961c7be98ce3ad9213d308e96397d53e8588e90926506dddd
This work supplies all eight patients' numerical inputs, pulse roles, metadata, task selection, source hashes, and extraction code. We retain those tasks and add correct-site coverage, acceptance of other sites, and response-set size.
usesdoi:10.18112/openneuro.ds004696.v1.0.1
Source SEEG data from OpenNeuro ds004696 v1.0.1; the cached exact-version metadata identifies this release as CC0.
doi:10.1523/JNEUROSCI.2201-22.2023

Cited by

2 publications

Public catalog entries on Sciscape.

supportsCovariance rank checks fitted to five reference SEEG vectors and one query accepted every query in rank-five groups at 83 sites in eight people
This work shows that response acceptance must be reported with rejection of different signals. Our separate covariance test shows complete acceptance when six centered vectors have rank five. The readouts, features, and comparison differ.
usesThree-nearest response scoring increased the site acceptance gap by 1.43 percentage points in an eight-patient SEEG test
This work uses the exact eight-patient feature inputs, pulse maps, and metadata from this publication. It compares a fixed three-neighbor score with the mean-distance score under a past-reference time rule.