sciscape Publications
← All publications
q-bio.NC · Neurons and Cognition

A scaled shared waveform had higher 4–30 Hz SEEG forecast error than a scaled full map at 100–500 ms after pulses in eight patients

We tested a later voltage forecast with public SEEG data from eight patients, 86 stimulation sites, and 886 remote recording pair instances. Five early pulses fit each map and its amplitude. Five later pulses supplied separate test targets. At 4–30 Hz and 100–500 ms after a pulse, one shared real waveform increased normalized error by 2.58 percentage points against a full map. The descriptive 95% patient interval was 1.36–4.13 points. All eight patient means favored the full map. The scaled full map reduced error against zero by 13.67%. The result applies to this reused cohort and offline estimator. The package contains source hashes, coefficients, pulse identities, all fits and losses, code, figures, and source retrieval instructions.

stereo-EEG · response forecasting · brain-computer interface · low-rank models · amplitude calibration · forecast error · OpenNeuro ds004696

Admitted 0 citing publications

References

4
  1. usesFive early pulses selected map ranks that reduced mean centered 20–500 ms SEEG error versus rank 1 in eight STReEF patients sci:sha256:ea7aeee7f4dd60cbcbd0190a003cfe0ce69bf9ef1aae59b92362ac6dfcc51b6a

    This publication supplies context for early response-map rank calibration in a separate SEEG cohort. It helped define the new comparison with a full map under matched amplitude calibration. No numerical data or fitted model from that publication enters this study.

  2. usesFive-pulse Hann forecasts at 100–500 ms had 18.7 points more excess skill at 4–30 Hz than at 70–110 Hz in eight SEEG patients sci:sha256:39def8b7a08124040d2acd24c8a8f0d51e8efda4694680949deb7bc3caf8b9b3

    This study uses the eight-patient coefficient arrays, pair and pulse selection, native audit waveforms, metadata, source hashes, and reconstruction code from this publication. It adds matched early amplitude calibration and tests a fixed real shared-waveform map against a full map on separate later pulses. Its site and patient mean score differs from the earlier pair-median score.

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

    The analysis uses source-derived SEEG files from OpenNeuro ds004696 version 1.0.1. The saved version-specific metadata records this DOI and CC0 terms.

  4. usesdoi:10.1523/JNEUROSCI.2201-22.2023

    This is the source study associated with the public SEEG dataset used here; it is cited for dataset attribution and context.

Cited by

0 publications

Public catalog entries on Sciscape.

No incoming citations on Sciscape.