Removal 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
We tested a short linear relation between recorded room sound and eight bipolar SEEG pairs in each of nine people. The data were OpenNeuro ds004703, version 1.1.0. We fitted a 0–10 ms waveform map on 54 early passages. We subtracted its predicted component and tested fixed 400 ms sound-amplitude readers on 100 later passage records. Mean MSE skill was 0.1385 before subtraction and 0.1406 after subtraction. Skill before minus skill after was -0.00214, with a descriptive 95% person bootstrap interval of -0.00730 to +0.00341. The exact sign-flip p value was 0.5430. Mean correlation was 0.3429 before and 0.3347 after. One person had a larger correlation loss, from 0.2293 to 0.1331. A known added zero-lag room waveform lost 99.988% of its energy on average. All 154 fresh EDF range hashes matched the cited source. Raw reader predictions matched the parent exactly. A full offline replay passed. The SCI edition retains fixed models, code, provenance, results, and figures; source recordings, audio, metadata, and derived input caches are linked but omitted. The tested subtraction caused little change in mean skill. It does not rule out other contamination or prove neural signal origin. This is a new subtraction analysis of previously tested listening records; it does not test brain stimulation. Source terms conflict: the saved dataset description says CC0, while its README states noncommercial and no-identification restrictions.
stereoelectroencephalography · SEEG · speech · brain-computer interface · acoustic contamination · linear regression · signal processing · MSE skill
References
4- usesFixed SEEG readers had higher mean sound amplitude correlation at 400 ms than at zero delay on 100 unused later passage records in nine people sci:sha256:3c8b962f91454c5d1b32469fb10a6ecb9534841b65f6b7ab946bc56a062e2126
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.
- 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
The inherited EDF header parser and adjacent-contact rules use this earlier MIT source code. Its notices are preserved.
- 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 study keeps the eight-pair combined-feature 400 ms readers and all their parameters fixed.
- usesdoi:10.18112/openneuro.ds004703.v1.1.0
The analysis used OpenNeuro ds004703 version 1.1.0, including the recorded room channel, SEEG, task timing, and source stimulus records.
Cited by
0 publicationsPublic catalog entries on Sciscape.
No incoming citations on Sciscape.