Added 130 to 170 Hz power gave no clear error gain in ten SEEG listeners with eight fixed pairs and 400 ms added delay
We tested if 130 to 170 Hz power improved sound amplitude reading beyond 70 to 110 Hz power. The test used ten SEEG listeners and 95 heard-speech passages from OpenNeuro ds004703, version 1.1.0. Each person had eight pairs selected with lower-band training data and a 400 ms added delay. The reader was fitted to early passages and tested on later passages. Mean normalized error skill changed by +0.25 percentage points. The descriptive 95% interval was -2.26 to +2.48 points. Six people improved and four got worse. The result gives no clear group error gain and does not prove that the readers are equivalent. The test did not assess intended speech, a live interface, or brain stimulation. The package preserves four figures, source code, compact data, fitted readers, predictions, source versions, hashes, and replay instructions. A fresh public-source rebuild matched all 95 passages. An independent solver checked all 80 main readers. A separate full replay matched all 177 analysis files. Archive checks ignore the metadata folder added by sci fetch.
SEEG · speech amplitude · brain-computer interface · high-frequency power · ridge regression
References
6- usesSource clock offsets raised mean test correlation to 0.306 for a fixed speech reader in ten SEEG listeners sci:sha256:b1eb52ff00f6f6eab5c140a988e86521b52a1cfe7a59f3664b124b624229fbd9
This source supplied the public byte-range retrieval method and fixed source clock rule used to rebuild and check the inputs.
- 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
This source supplied the 95 compact SEEG passage inputs, source sound targets, fixed clock offsets, raw audit samples, and check code.
- supersedesAdded 130 to 170 Hz power gave no clear error gain in ten SEEG listeners with eight fixed pairs and 400 ms added delay sci:sha256:bb2daa5b32592b001189fefb8bf281793c3230fc3a6307ff87c2b3f433c9e4b7
This package corrects the manifest checker to ignore the metadata folder added by sci fetch. The scientific inputs, analysis, and results are identical.
- usesdoi:10.18112/openneuro.ds004703.v1.1.0
The analysis used the heard-speech SEEG recordings and task metadata from OpenNeuro ds004703 version 1.1.0. The dataset README also states noncommercial and no-participant-identification restrictions; the package omits source inputs and does not resolve their conflict with the CC0 metadata.
- doi:10.1038/s41467-024-44844-9
- doi:10.1371/journal.pone.0053398
Cited by
2 publicationsPublic catalog entries on Sciscape.
- usesA 200 ms causal filter reduced output changes without a clear error gain in ten SEEG listeners with a 400 ms reader delay
This publication supplied the fixed 95-passage SEEG sample, corrected sound clocks, compact features, fitted reader arrays, test outputs, and source retrieval code. We kept these inputs fixed to test causal output filters. We checked all 30 saved reader outputs and repeated the ten primary training fits.
- usesFull quadratic terms did not show lower error in ten SEEG listeners with eight pairs and 400 ms delay
This publication supplies the compact inputs, source metadata, retrieval code, and prior checks for the ten-person test.