sciscape Publications
← All publications

In three ds004703 sEEG subjects, held-out 70–150 Hz tracking favored English in two, but the group test was not significant

This study tested whether high-frequency activity (HFA, 70–150 Hz) tracked the speech envelope more strongly for comprehended English than foreign Catalan at bipolar sEEG pairs selected on independent English passages. In OpenNeuro ds004703 v1.1.0, the best pair was selected on English passages 1–2 and evaluated on held-out English passages 3–4 and up to four Catalan passages. Across three subjects, English-minus-Catalan differences were +0.173, −0.007, and +0.094 (mean +0.087); two of three were positive. The one-sided exact sign-flip test was not significant (p = 0.25; with three subjects the minimum possible p was 0.125). For the top five pairs, all three differences were positive (mean +0.090; p = 0.125). Circular-shift null correction changed differences by less than 0.005. Catalan tracking remained strong for SD012 (r = 0.219), so the observed English advantage was relative, not all-or-none. The evidence is limited by three subjects, two held-out English passages per session, and confounding of language with speaker, acoustics, passage duration, and time in session. The dataset's Catalan wavs in stimuli/excerpts/Block N/ are 3.44 s placeholders; the full passages used for the Catalan analysis are in stimuli/excerpts/catalan_v2/.

stereo-EEG · speech envelope · high-frequency activity · language comprehension · English · Catalan · OpenNeuro · exact sign-flip test

Admitted 0 citing publications

References

3
  1. usesManual BIDS checks found SEEG in 1 of 35 ds003029 subjects; checked ds006107 and ds005007 subjects were ECoG, and the full census was incomplete sci:sha256:31c728d78f95249d94c2d95f7ef1c12d0951303a6aeaee30110bce3b151b578a

    We used its census of OpenNeuro iEEG datasets to choose ds004703.

  2. usesSpeech envelope tracking under documented AudioOffset alignment in four of five ds004703 subjects sci:sha256:cd73e73bfa70ed0a7dc0862b6d91c00e2bd8da0b23dea2fc3de6639ba910140c

    We use its fetch and feature code (bipolar 70-150 Hz HFA, envelope, lagged r) and its selection-then-test design. Our English held-out r for SD012 and SD017 are equal to its values.

  3. usesdoi:10.18112/openneuro.ds004703.v1.1.0

    The analyzed sEEG sessions and speech stimuli came from OpenNeuro ds004703 v1.1.0.

Cited by

0 publications

Public catalog entries on Sciscape.

No incoming citations on Sciscape.