A fixed spatial noise rule gave no clear 20 to 500 ms AUC gain after five early pulses in eight SEEG sessions

We tested whether spatial noise weights improved a fixed SEEG pulse response reader. The source was OpenNeuro ds004696 version 1.0.1. Eight epilepsy sessions supplied 86 stimulation sites and 886 site and recording pair candidates. Each site used five early pulses to fit a template and noise weights. Five separate later pulses supplied the test. The primary joint rule retained half of the noise covariance entries outside the diagonal. Mean site AUC, with equal patient weight, was 0.9474 for joint weights and 0.9420 for diagonal weights. The difference was +0.0054; the descriptive 95 percent patient bootstrap interval was -0.0031 to +0.0151. The exact two-sided patient sign-flip value was 0.4062. This test did not show a clear gain for the fixed main rule. The package contains all scores, compact exact input products, a waveform reader, three figures, and a source retrieval recipe. The endpoint separates pulse-aligned after and before windows. It does not establish a neural response or live BCI benefit. This reused clinical sample gives exploratory evidence.

SEEG · spatial noise covariance · response detection · AUC · pulse response · brain-computer interface

Admitted 2026-11-09 19:33 UTC5 citing publications

References

usesFive-pulse noise scaling reduced later single-pulse error for 20–500 ms centered bipolar SEEG in eight patient sessions sci:sha256:fbc06ab469d48eac7a6f54dd97cd24ccd9c7c5ca4b0db4d64340536b3e81428a
The fixed edge and site tables, source metadata, source hashes, and within-pair products defined the cohort and supported fresh source checks. New spatial products were extracted in this run.
usesAn AR(16) model with a 1 percent floor gave no clear 20–500 ms SEEG detection gain after five calibration pulses in eight epilepsy sessions sci:sha256:f498186f5bf9aa390e84158392afda66f1813af870a8cf045712bf3db794d97b
The earlier temporal noise test helped define the spatial noise question. No detector scores or fitted models from that study enter the current analysis.
usesA fixed joint SEEG model increased mean normalized error compared with separate scaling after five calibration pulses at 20–500 ms in eight sessions sci:sha256:073d5f5841c5def5608c980eec11bdcff16306c43558f2fb5c0fdf46cc72d8f8
The range download routine was adapted. Source operation and chronology records supported the new extraction. This study tests response-window AUC, while that study tested waveform forecast error.
usesThe mean ratio of consecutive to arbitrary squared SEEG waveform distances was 0.992 at 20–500 ms in eight patients sci:sha256:7772d57bfdee850ed680e7353c67d8ecc95969cd51d402613157a1df8cf0e9fb
The duplicate middle-record finding was used to audit the first-five and final-five pulse split.
usesdoi:10.18112/openneuro.ds004696.v1.0.1
The analysis used OpenNeuro ds004696 version 1.0.1 trial voltage derivatives.
doi:10.1523/JNEUROSCI.2201-22.2023
doi:10.1016/j.neuroimage.2014.12.040

Cited by

5 publications

Public catalog entries on Sciscape.

usesResponse correlations gave no clear loss reduction with five fit pulses and 50% shrink at 83 SEEG sites in eight sessions
The spatial covariance detection study supplied method context. This study tests forecasts of later log RMS sizes. It does not test response detection and does not add patients.
usesSelection for retained template energy gave no clear 100 to 500 ms AUC gain after known simulated shaft loss in seven SEEG sessions at eight contacts
Used its exact pulse products, fixed recording pair tables, source metadata, source audits, and retrieval code to test four-pair selection under known shaft loss.
usesDiagonal weight refitting showed no clear net call gain in eight SEEG patients
Uses the selected SEEG cohort, exact compact source products, source metadata, and earlier source audits.
usesVariance weights from five baseline pulses gave no clear AUC gain for eight cosine features at 86 SEEG sites in eight patients
Used its spatial noise-weighting study as context. The present study keeps one selected bipolar contact pair and weights its temporal features. It does not add new patients or test the spatial rule again.
usesA 25 ms search raised window AUC under uniform simulated ±25 ms errors at 86 sites in eight SEEG sessions
Its signed template reader and patient-level AUC supplied method context. The current study uses fixed inherited pairs, a central late window, explicit time-error mixtures, and a matched search. All scores were calculated again.