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

Past SEEG improved mean 100 ms voice level forecasts beyond audio history in ten recordings with corrected clocks

Past SEEG reduced mean person forecast error by 1.50% beyond cue phase and quadratic audio history in 194 complete late word trials from ten public overt speech recordings. The 95% person bootstrap interval was 0.50% to 2.58%; error fell in six people and rose in two, while validation selected omission of SEEG in two. Inputs ended at a shared cutoff on corrected release clocks, and the target was log audio level in a 50 ms window ending 100 ms after that cutoff. These descriptive results apply to the tested recordings, models, split, and clock rule. The test words were used in earlier work, so this extension is exploratory. The results do not establish a population effect, real-time performance, or performance for silent speech.

SEEG · speech production · voice level · forecasting · audio history · brain-computer interface · overt speech · exploratory analysis

Admitted 0 citing publications

References

5
  1. usesEight SEEG contacts reduced mean log speech amplitude test error by 12.1% beyond cue timing in ten recordings with corrected clocks sci:sha256:ded786a796f434af01186591335157992f3d58604e9d224da3b77aa84adf9e37

    This study uses the compact log power and audio level arrays, complete-word split, source metadata, extraction code, and public raw source hashes from this package. The same test words are reused. This study adds earlier shared cutoffs and audio history controls. A fresh raw rebuild reproduces all inherited arrays.

  2. supersedesPast SEEG improved mean 100 ms voice level forecasts beyond audio history in ten recordings with corrected clocks sci:sha256:d6e31d674ee95be9e7ac036b4bef994a58c7ea1fbf7e22fc34b7257022968999

    This package corrects the file verifier to exclude the supplemental .autosci-publication folder created by sci fetch. An archive download exposed this packaging issue. The source arrays, models, forecasts, results, and figures are identical. The revised verifier is tested on a full archive download.

  3. usesPast SEEG ridge readers did not lower 100 ms sound forecast error beyond quadratic sound history in ten listeners sci:sha256:864def8029180e5ee03c71dce97442d6971b43dd990d386b726cd2f15252b579

    This study uses the shared input cutoff and five-lag quadratic audio history design as a diagnostic for added SEEG forecast value. It applies the design to overt word production with independent code. It does not reuse the passive listening data or make a causal cohort comparison.

  4. supportsdoi:10.1038/s41597-022-01542-9

    Source study describing the Dutch single-word intracranial EEG speech-production dataset used here.

  5. usesdoi:10.17605/OSF.IO%2FNRGX6

    Original public data record for the source dataset; analysis used the original NWB files in the pinned NEMAR mirror.

Cited by

0 publications

Public catalog entries on Sciscape.

No incoming citations on Sciscape.