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

In 16 SEEG patients, HFB speech–music counts were not above chance, while alpha and beta speech > music counts were

Across 16 patients with SEEG contacts in OpenNeuro ds003688 v1.0.7, we tested 579 adjacent good-contact bipolar pairs during 13 alternating speech and music film blocks. For each block, we analyzed seconds 2–7 and tested speech–music differences in log10 band power using exact block-label permutations; group counts used a subject-shared-label null. The primary high-frequency broadband (70–150 Hz) tests were not significant: 26 speech > music pairs (null mean 14.2; p = 0.058; Holm p = 0.116) and 14 music > speech pairs (p = 0.47). Secondary counts were above chance for alpha speech > music (43 pairs; Holm p = 0.011), beta speech > music (37; Holm p = 0.026), and theta music > speech (37; Holm p = 0.048). An exploratory per-subject direction test was not significant in any band; a few subjects with many pairs contributed most of the excess. Per-pair differences agreed across non-overlapping early and late windows of the same blocks (HFB mean Spearman rho 0.53 in 14/15 subjects, p = 0.0001), indicating repeatable block differences but not category selectivity. Speech and music blocks also differed in picture, sound level, and attention; only 5 seconds per block were analyzed, and the permutation assumes exchangeable block labels. The results therefore do not establish speech- or music-specific responses.

SEEG · speech · music · film · band power · high-frequency broadband · OpenNeuro · block permutation

Admitted 1 citing publication

References

3
  1. usesHigh-gamma speech–music differences did not repeat across film halves in SEEG depth contacts sci:sha256:6175fbed365033e76975730ad326cf43bff4af2df29f1de1d22c9f96d39f4bf5

    We used its archived per-pair results (results/pairs.tsv, seconds 10-20 of each block, SHA-256 76b41f70...) in exploratory test E2. The per-pair speech-minus-music differences of our seconds 2-7 window agreed with it (HFB mean rho 0.53, 14/15 subjects; beta 0.49). Our early-window HFB count null and low-frequency count excess agree with its HG split-half null and 8-30 Hz split-half repeat.

  2. supportsIn 2 of 5 SEEG patients, significant HFB differences favored music over speech in film blocks sci:sha256:02758c50816f528414dacf219e42e90906b45fa6be431490011c6938a87e136d

    With all 16 SEEG subjects and an earlier window (seconds 2-7), the HFB count test also did not show a speech-specific HFB increase, and sub-09 again had music > speech HFB pairs (left TOL lead; it reported left parietal MPL8-9). It differs in one point: we found more 8-30 Hz speech > music pairs than chance, but it found no patient-level 8-30 Hz effect in 5 pure-SEEG patients.

  3. usesdoi:10.18112/openneuro.ds003688.v1.0.7

    The analysis used the film-task SEEG recordings from OpenNeuro ds003688 version 1.0.7; this is the recorded version-specific dataset DOI.

Cited by

1 publication

Public catalog entries on Sciscape.