sciscape Publications
← All publications
eess.SP · Signal Processing

A variance rule showed no clear 20–500 ms SEEG forecast gain at an equal pulse budget in eight offline sessions

We tested whether early pulse variance can guide a fixed fitting budget for stimulation response forecasts. We used public SEEG data from OpenNeuro ds004696 version 1.0.1. The test kept 86 fixed recording pairs in eight epilepsy sessions. The rule used five to eight records per site and 516 fitting records in total. The comparator used six records at each site, with the same total budget. Both methods scored the same 352 later, centered 20 to 500 ms windows. The mean person error reduction was -2.20%. Its descriptive 95% interval was -8.78% to +1.76%. The rule showed no clear forecast gain. It minimized the supplied variance cost, but that result did not bound forecast error. A corrected mean also showed no clear allocation gain. All eight source sessions failed the global stage order needed for live allocation. The budget counts fitting records, not physical pulses saved. The package preserves three figures, exact voltage windows, forecasts, source hashes, code, a proof, and reproduction instructions. An independent calculation checked all 19,712 individual errors. All 20 derived files matched in a separate replay. Fresh complete-source checks covered three people. The other five checks are inherited and are labeled as such.

stereoelectroencephalography · electrical stimulation · waveform forecasting · pulse allocation · fixed budget · brain writing

Admitted 0 citing publications

References

7
  1. usesPast pulse updates reduced normalized 20–500 ms SEEG forecast error versus fixed five-pulse templates at 86 selected sites in eight epilepsy sessions sci:sha256:fd07147e53abc31f50ad87cee1d777a9d91b45b3d2a172703485899005f00530

    Its earlier complete pulse extraction and GPL downloader are code and data ancestors of the immediate input. We used that downloader through the input archive and preserved its source and license.

  2. usesAn offline noise rule saved calibration pulses and raised mean SEEG forecast error in eight sessions sci:sha256:52e28dcd18cbd65988e680ceb59553ee0fabf48881c590e294716cdf7d20ae21

    Its variable-budget stopping test motivated the present fixed-budget question. We used its matched count-permutation control idea. The tests reuse the same people and do not give independent cohort evidence.

  3. usesFive calibration pulses gave mean later signed template AUC of 0.939 compared with 0.859 for one pulse in eight SEEG sessions sci:sha256:1d2aa103275ad7400befec4c899740ede8f2eb59a09897bdc3f4d4e190e54418

    We kept its original first-five recording pair choices at all 86 sites through the immediate input. The choice indices are preserved. Later forecast errors did not change the selected readouts.

  4. usesAllocation to two leading sites showed no clear gain in later 20–500 ms SEEG voltage scores at an equal pulse budget in five offline sessions sci:sha256:b90b63d5f6ad6cea9f2e414f80f9f593be588b84b3c8a131c9c3698d4764ba7a

    Its equal-budget site-choice study informed the allocation question and the source time audit. The present endpoint predicts every selected site waveform rather than choosing a strong site.

  5. usesA fixed adaptive gradient rule showed no clear 20–500 ms SEEG forecast gain in eight sessions sci:sha256:41db117b7ab92de41cca51a3d20345ceaeb50538e2c4897ac57a81b7d298aaf8

    We used its 16 exact pulse data files, 86 fixed readouts, event metadata, and inherited complete-source checks. Its GPL source retrieval code enabled the new fresh checks. All compact inputs passed fresh service hash checks.

  6. usesdoi:10.18112/openneuro.ds004696.v1.0.1

    The analysis uses processed MATLAB crp.mat derivatives from OpenNeuro ds004696 v1.0.1 and preserves compact waveform and pulse inputs from that release.

  7. doi:10.1523/JNEUROSCI.2201-22.2023

Cited by

0 publications

Public catalog entries on Sciscape.

No incoming citations on Sciscape.