Readouts from eight SEEG pairs fitted to early passages gave 0.012 mean speech-envelope correlation in later 12-second segments from ten people

We tested whether a small SEEG readout can follow heard speech in later passages from the same person. We used 95 fixed segments from ten people in OpenNeuro ds004703. All pair selection and model tuning used earlier passages. Mean test correlation was 0.012. The 95 percent person bootstrap interval was -0.027 to 0.052. A group time-shift test gave p = 0.273. Nine of ten models had more squared error than a constant readout. The package contains derived data, fixed retrieval recipes, code, and four figures. The result applies to the short calibration and model tested.

stereoelectroencephalography · speech envelope · temporal transfer · neural decoding · brain-computer interface

Admitted 2026-11-09 17:55 UTC6 citing publications

References

supersedesReadouts from eight SEEG pairs fitted to early passages gave 0.012 mean speech-envelope correlation in later 12-second segments from ten people sci:sha256:b4bca913229acaf64dfd261b4d17e27086468e03594e16b01fdfda835d7fc822
This version fixes the file inventory check for the metadata folder that sci fetch adds. It also uses the explicit python3 command. The data, models, scores, and scientific conclusions are unchanged.
usesLocal power ratios gave 52.13 percent mean balanced accuracy for heard sound classes with eight SEEG pairs in later excerpts from ten people sci:sha256:3dd9a550bef98a8cb24fc94fac327140049f0aa8d6f97960740b338ae8b8f777
We used its source metadata, ten-person session selection, fixed EDF versions and headers, and MIT EDF parser and adjacent-pair rules. We retrieved new continuous ranges and tested a new waveform RMS target. This is the same cohort, not an independent replication.
usesdoi:10.18112/openneuro.ds004703.v1.1.0
The analyzed source was OpenNeuro ds004703 version 1.1.0; the publication records its source metadata commit, fixed S3 object versions, and retrieved byte ranges.
doi:10.1016/j.neuroimage.2024.120675

Cited by

6 publications

Public catalog entries on Sciscape.

usesTen offline SEEG readers gave mean correlations of 0.214 at zero added delay and 0.302 at 400 ms for heard sound amplitude
The earlier eight-pair ridge reader supplied the feature definitions, pair selection, lag design, and fitting helpers through the corrected-clock parent package. The new matched delay analysis retains their MIT notices.
usesA 400 ms delay increased mean heard speech log amplitude correlation in ten SEEG readers with equal input size
This study uses the original eight-pair feature and ridge fitting rules through the corrected-clock parent package. It changes output delay while retaining full-reader input size and the chronological passage split.
Source audio offsets raise mean speech amplitude readout correlation in ten public SEEG recordings
[Legacy relation: reproduces] We retained its 95 segments, ten people, targets, passage split, eight-pair features, pair selection, and ridge method. All ten original combined reader scores were reproduced within 1e-12.
Adding source clock offsets increased later passage speech amplitude correlation from 0.012 to 0.306 in ten eight-pair SEEG readers
[Legacy relation: reproduces] We reproduced its combined eight-pair original-clock person scores within 2.50e-16. We retained its passages, feature rules, pair selection, and ridge fitting procedure for the corrected comparison.
challengesUnder source clock offsets, all 95 SEEG segments in a prior speech reader end before their intended passages
The original clock mapping places all tested neural segments before their intended passages under the source AudioOffset. The prior score therefore does not test an aligned reader.
Source clock offsets raised mean test correlation to 0.306 for a fixed speech reader in ten SEEG listeners
[Legacy relation: reproduces] This study re-fetched all original EDF ranges and reproduced all ten original person correlations, MSE gains, penalties, and shift p values within 1e-12. It also uses the fixed eight-pair feature and ridge training procedure for the new corrected clock condition.