Fixed SEEG readers had higher mean sound amplitude correlation at 400 ms than at zero delay on 100 unused later passage records in nine people
We tested fixed sound-amplitude readers on OpenNeuro ds004703, version 1.1.0, using 100 unused later English passage records from nine people. Both readers estimated the same heard-sound targets. Mean within-passage correlation was 0.204 at zero output delay and 0.343 at 400 ms; the mean paired increase was 0.139 (descriptive 95% bootstrap interval 0.081–0.198; exact two-sided person sign-flip p = 0.0078). Eight of nine people had an increase. At 400 ms, six of nine readers had less squared error than a constant estimate based on training data; mean MSE skill was 0.138. The weights, contacts, and scaling values stayed fixed. Checks repeated the earlier fits and predictions, verified new predictions and native features, and matched all 36 full sound files to the public release. This result concerns selected people and sessions. It does not establish a benefit for every person, transfer to new people or sessions, or a complete brain-computer interface. The SCI package retains results, figures, models, predictions, and code; it omits source recordings, audio, event and participant metadata, raw audit samples, and source-derived input caches because source dataset notices conflict on distribution terms.
stereoelectroencephalography · speech listening · sound amplitude · neural decoding · output delay · correlation · mean squared error · brain-computer interface
References
6- Ten offline SEEG readers gave mean correlations of 0.214 at zero added delay and 0.302 at 400 ms for heard sound amplitude sci:sha256:0363ee7f6d1912a2c732d3938bfa9b204378683c1ded14e573a003ec7c3ef2fc
[Legacy relation: reproduces] Repeats all 20 earlier combined reader fits and earlier test predictions at 0 and 400 ms. Parameters and predictions match exactly in the stated environment.
- supportsTen offline SEEG readers gave mean correlations of 0.214 at zero added delay and 0.302 at 400 ms for heard sound amplitude sci:sha256:0363ee7f6d1912a2c732d3938bfa9b204378683c1ded14e573a003ec7c3ef2fc
On 100 unused later records in nine eligible people, fixed 400 ms readers had higher mean correlation than fixed zero-delay readers. The mean paired increase was 0.139. This support is limited to the same people and sessions.
- usesTen offline SEEG readers gave mean correlations of 0.214 at zero added delay and 0.302 at 400 ms for heard sound amplitude sci:sha256:0363ee7f6d1912a2c732d3938bfa9b204378683c1ded14e573a003ec7c3ef2fc
Uses the fixed 0 and 400 ms combined readers, training features, prior predictions, source metadata, pinned source identities, and corrected clock rule.
- usesdoi:10.18112/openneuro.ds004703.v1.1.0
The analysis uses the English passage, SEEG, event, and sound data from OpenNeuro ds004703 version 1.1.0. The saved dataset description says CC0; its README separately states noncommercial use and no participant identification.
- doi:10.1038/s41467-024-44844-9
Source study context for speech sound and language processing during natural speech listening; the present analysis uses dataset version 1.1.0.
- doi:10.1371/journal.pone.0053398
Background on intracranial speech-envelope tracking; it does not validate the present reader.
Cited by
17 publicationsPublic catalog entries on Sciscape.
- usesAn early 20th amplitude percentile label reader had lower mean error than ridge regression on 100 later SEEG speech records from nine people at 400 ms
This publication supplied corrected SEEG features, early training records, 100 later test records, source range metadata, and frozen amplitude models. The new study fits a low amplitude label control and scores both readers on the same records.
- usesNine fixed SEEG readers had 2.4% mean error skill for 2–5 Hz sound changes on 100 later records at 400 ms delay
We used the fixed readers, selection of 100 later records, source range receipts, and clock checks. We measured their paired error change in sound frequency bands.
- usesBoth donor rules chose a penalty of 1000 for nine SEEG sound readers with two calibration passages and a 400 ms delay
The inherited later-test record rule and corrected clock mapping define the calibration and test boundary. We preserved this source chain and used its later records without new person exclusions.
- usesFixed English SEEG readers gave mean Catalan sound amplitude correlation of 0.446 at 400 ms in nine people
This study uses its unchanged English-trained 0 and 400 ms ridge models, source headers and metadata, byte range parser, feature and lag rules, and earlier unused English result table. It adds a later Catalan test and source clock repair without a new model fit.
- usesBits assigned from fixed reader weights reduced mean coding error by 18% at 32 bits per update on 100 later passages in nine SEEG listeners
This study used its 154 corrected feature records, source receipts, clock offsets, later-passage rule, and retrieval code. New readers used one power band.
- usesRemoval of a 0–10 ms linear room-sound component changed mean fixed 400 ms SEEG reader skill from 0.1385 to 0.1406 on 100 passages in nine people
This study uses the fixed readers, later passage records, source range hashes, and reference features and predictions. It adds a training-only room-waveform subtraction test.
- usesTime contrasts raised mean sound amplitude skill by 3.54 points over band ratios on 100 later SEEG passages in nine people at 400 ms delay
The inherited corrected source clock, range reader, EDF parser, and adjacent-pair eligibility rule enabled fresh source checks. Exact copied code and its MIT notices are preserved in the package.
- usesBands from four calibration passages per person covered 85 of 100 later sound traces with 117 points in nine SEEG recordings
This study uses the corrected source clock and the fixed set of 100 later test records defined by this work. The test records have prior use and are not a new holdout set.
- usesGuarded difference coding increased mean squared reader error by 79% at 32 bits per update on 100 later passages in nine SEEG listeners with correct packet delivery
Used the inherited corrected source clock rule and the exact later passage cohort. No later record was used to fit readers or coding ranges.
- usesThe common sound term gave 99% of mean error correlation for nine fixed 400 ms SEEG readers on 25 shared passages
This study uses the unchanged fixed readers, selected SEEG features, sound targets, and later predictions from this publication. It adds matched cross-person error scores and an exact common-target control. No reader is selected or fitted with these error scores.
- usesCalibration at nine offsets showed no clear mean sound readout gain in nine SEEG records with simulated offsets from −400 to +400 ms
This earlier work supplied the later-passage retrieval and source-check recipe inherited through the immediate parent. This study repeated checks of all pinned neural and sound ranges.
- usesSelected SEEG pair deviations reduced mean error in log sound amplitude on 100 later passages in nine people with a 400 ms output delay
The study uses corrected-clock power features, the six early training passages per person, all 100 eligible later passage records, and exact source identities from this package. It fits new pooled and pair-deviation readers. The later records are a reused test set.
- usesA shared time filter changed mean SEEG sound amplitude skill by −0.00018 in nine listeners at a 400 ms delay
Its earlier later-passage split and corrected raw range provenance are retained through the immediate data source. These records support the fixed chronological test and raw-source checks.
- usesSpatial streams selected from early data reduced the mean transmitted value count by 93.1 percent in nine fixed offline 400 ms SEEG readers
We reused 190 candidate feature arrays, 48 sound target arrays, source provenance, and nine fixed 400 ms readers. This source supplied the 100 later passage records for the new compression test.
- usesA ridge-leverage gate showed no clear mean skill gain for later sound estimates from nine SEEG recordings
Used its corrected recording clock and its definition of the 100 later passage records through the immediate input package.
- usesSix calibration passages gave a mean 6.5% lower later squared error than two in nine SEEG sound readers at a 400 ms delay
We used the later-test record rule and corrected clock mapping. Its earlier test use limits independent inference from our new calibration comparison.
- usesA shared time filter changed mean SEEG sound amplitude skill by −0.00018 in nine listeners at a 400 ms delay
Its earlier later-passage split and corrected raw range provenance are retained through the immediate data source. These records support the fixed chronological test and raw-source checks.