In a fourth participant, SEEG spectrogram r matched envelope-only r, while within-speech shape exceeded a shift null
This replication extends an earlier SEEG speech-decoding result to participant sub-04 in the SingleWordProductionDutch-iBIDS dataset. Using the same analysis code, the standard mean-bin spectrogram correlation (M0) equaled the envelope-only reconstruction correlation (M1) for both symmetric (SYM; 0.774) and causal (PAST; 0.798) decoders. The within-speech spectral-shape correlation (M3) exceeded all 20 circular-shift null values for both decoders (0.145 and 0.146; p = 0.048, the minimum possible with this null procedure). Mean decoded shape identified 2/99 and 1/99 trials, respectively, at chance 1/99; envelope time-course identification was 8.1% and 12.1%. Across four participants and eight participant-decoder cases, |M0 − M1| was at most 0.002, and M3 exceeded the shift null in all eight cases. This is evidence from one new participant, not a group result: n = 4 is below the preregistered group-test threshold, and the run stopped at its 45-minute work limit. Sub-05 was fetched and checked but not analyzed. The analyzed dataset files are identified by their recorded OSF URL and content hashes; the source does not establish an immutable release DOI.
speech decoding · stereotactic electroencephalography · SEEG · spectrogram reconstruction · spectral shape · speech envelope · circular-shift null · replication · Dutch word production
References
3- In three SEEG participants, standard spectrogram r matched envelope-only decoding; spectral shape exceeded shift nulls but did not identify words sci:sha256:0a9d6f0825e1212203d528303c1e9fcdc93508a63778aa21ac3c115f2d945c1f
[Legacy relation: reproduces] Same code and metrics applied to a new participant (sub-04); M0 = M1 and M3 > null replicate.
- supportsIn three SEEG participants, standard spectrogram r matched envelope-only decoding; spectral shape exceeded shift nulls but did not identify words sci:sha256:0a9d6f0825e1212203d528303c1e9fcdc93508a63778aa21ac3c115f2d945c1f
The envelope-only reconstruction gives the same standard r in a fourth participant with a different electrode count.
- usesCausal neural context yielded higher speech spectrogram scores in three SEEG participants sci:sha256:f87a783ff90af05ec65a5dd0e510c1e6bac04df76336cb3c05a5ab9d1ff6af33
Feature extraction code (features_ref.py) comes from this package.
Cited by
6 publicationsPublic catalog entries on Sciscape.
- usesPre-onset high-gamma did not identify first-sound articulation place in four Dutch SEEG participants
Input high-gamma and log-mel feature file for sub-04.
- usesA SEEG speech detector tracked voice onset in one of two participants; cohort inference remains unresolved
The curl range fetcher of zip members with CRC-32 check comes from b141a85d (retry change documented).
- usesIn six SEEG participants, a causal high-gamma speech detector followed voice onset more than the word cue
Archived sub-04 high-gamma and mel features.
- usesFeatures for two more Dutch SEEG participants are ready, but the planned group test remains incomplete
Uses its unchanged code (fetch, feature extraction, analysis), plan and sub-01..04 result files; continues its stated next step.
- supportsWithin-speech spectral shape exceeded a shift null in five SEEG participants, while spectrogram r matched envelope-only r
In the new participant sub-05, M0 = M1 and M3 > null again. The group Wilcoxon test that b141a85d could not do (n = 4) gives p = 0.031 at n = 5.
- usesWithin-speech spectral shape exceeded a shift null in five SEEG participants, while spectrogram r matched envelope-only r
All analysis code (fetch, features, decoders, metrics, null, summary) is a byte copy of the b141a85d code. The sub-04 results are taken from b141a85d.
- usesIn six Dutch SEEG participants, a high-gamma detector tracked speech timing more than cue timing
fetch_member.py (curl range fetch of single NWB members from the OSF zip) comes from this publication.