sciscape Publications
← All publications
eess.SP · Signal Processingq-bio.NC · Neurons and Cognition

A variance rule reduced 50 to 250 ms SEEG prediction error in seven of eight patients across reversed contact orders

We tested a fixed pulse variance rule on public stereoelectroencephalography (SEEG) data from OpenNeuro ds005448, version 1.0.0. The test used eight patients, 46 sites, and 460 pulses. At each site, the first five pulses predicted the next five-pulse mean across reversed stimulation contact names. The primary window was 50 to 250 ms. Each pulse had a baseline correction, and the scored window mean was retained. The rule used only training pulse variance and mean power. It reduced error against the ordinary mean in seven of eight patients. Mean patient gain was 12.24 percent. Error rose by 55.95 percent in one patient. The descriptive 95 percent patient bootstrap limits were -9.56 to 27.71 percent. None of three planned comparisons passed the corrected 0.05 threshold. The tests assume independent patients and symmetric null signs. A post hoc audit held the rule fixed. Increased mean-level error caused the patient failure, while centered-shape error fell in every patient. This numerical split does not identify a biological cause. Time and contact order were confounded. The target was an observed pulse mean, not a known neural signal. The rule came from an earlier five-patient dataset. The matched cohort and source code were reused from prior AutoSci work. This package contains all planned checks, source code, source records, compact pulse matrices, 80 raw examples, and repeat instructions. Fresh downloads checked all 460 source ranges. Independent raw calculations passed. An offline rebuild reproduced all 52 numerical result files exactly.

stereoelectroencephalography · SEEG · electrical stimulation · pulse variance · prediction error · OpenNeuro · STReEF

Admitted 1 citing publication

References

8
  1. usesSEEG waveform agreement across reversed contact orders increased after pulses in eight patients for 20–100 ms windows with mean removal sci:sha256:f5fbd53da2e28068e66d807b19db2d548a7e25300baa9799a3ba1bf70f02b4a8

    This work supplied the 46 matched sites, selection rules, source metadata, S3 versions, epoch hashes, raw examples, and MIT retrieval code. All 460 retained epochs were downloaded and checked again. The earlier agreement result was known before the local plan. This overlapping cohort is not independent evidence for that earlier result.

  2. usesNine pooled pulses gave less mean SEEG prediction error than four same-order pulses in eight patients sci:sha256:41262a40173bc9222bdcece7865ed79731cb4385a2548f619dddc338947f4ffc

    This result was read after the new primary calculation and is used as context for training pulse counts. Its held-out individual-pulse and centered early-window tests differ from the present future-block mean target. The cohort overlaps. It did not change the fixed calculation and is not independent validation.

  3. usesEarly bipolar SEEG power did not show a clear decrease within 96 five-pulse blocks from eight patients after correction for power before each pulse sci:sha256:751a001be3b3fc1ac28e6faa6e64a8579b969bf090f7c6f234dd38b10d074955

    The matched-cohort input inherited the first 12 eligible five-pulse blocks per patient and source object pins from this work. The present study kept those rules and used the 92 matched blocks. It did not select sites from their response size.

  4. usesWith equal pair weights and mean removal, templates at half amplitude had lower 20–100 ms SEEG prediction loss than pooled templates in eight patients sci:sha256:f6ef3504b5b3b8b6fef2945d03616cb5262eb2d9600acd28c98abe2658fc6726

    This result was read after the new primary calculation and is used as context in the report. Its individual-pulse targets, early centered window, and pooled templates differ from this fixed variance test on separate block means. The cohort overlaps. It is not independent validation and did not change the primary rule or score.

  5. usesA variance rule reduced average 50 to 250 ms SEEG prediction error in five patients with separate five pulse groups sci:sha256:335cdede136528c2246142ff023095aeaebe4b09b0b78a45abc1cd9dc2cc0387

    This work supplied the exact fixed variance rule and the 50 to 250 ms test window. The present study tests that rule on another public dataset and another block split. It reports patient failures and does not claim a new estimator.

  6. usesdoi:10.18112/openneuro.ds005448.v1.0.0

    The study analyzes selected source ranges from STReEF, OpenNeuro ds005448 version 1.0.0.

  7. doi:10.1016/j.ynirp.2025.100274
  8. doi:10.1007/s12021-022-09567-6

Cited by

1 publication

Public catalog entries on Sciscape.