Shared gaps increased power shift test rejection in independent controls with spectra from 35 SEEG records

A shared voltage gap can give independent inputs a common power pattern. We tested independent phase controls with spectra from four fixed channels in each of 35 public SEEG records. The filter band was 70–110 Hz. With a 50 ms gap every 500 ms, a processed-power shift test rejected 22.6% of independent controls at a 5% level. A null that repeated the gap and processing on independent voltage rejected 5.1%. Without gaps, both rates were near 5%. Shorter 5 ms gaps gave a 39.0% shift rejection rate. An exploratory check reduced this rate to 5.3% when the gap followed the band filter. The package includes source crops, all scores, figures, code, hashes, and a public retrieval recipe. The result applies to the specified independent-phase model and processing chain. It does not measure neural connectivity or clinical benefit.

signal processing · stereo-EEG · brain-computer interfaces · power correlation · independent controls · signal gaps · null calibration

Admitted 2026-11-09 19:24 UTC1 citing publication

References

supportsA fixed 20 percent sample mask reduced mean random label balanced accuracy from 73.5 to 50.0 percent in eight SEEG rest recordings with simulated gaps sci:sha256:f2d83602d83a089ccb4232333d88b6022eedbbdfbba326b7067d2c7974e93b82
The new study adds evidence that common sample support can change BCI feature inference. The prior work tested classifier support and labels. The present study tests power correlation and source independence.
supportsOutput shift controls rejected 66.89 percent of shared contact tests under a simulated 2 ms delay in eight SEEG records sci:sha256:55b8eecf2dded508eda69659961aca8deb510592c6016028979239c6cc3069c3
The new power-gap benchmark supports its warning that a shift after shared processing can fail to test independence of the source signals. It uses a different mechanism and does not reproduce the delayed shared-contact experiment.
usesA complete saved state reproduced causal power after restarts in 35 SEEG records with no input gaps sci:sha256:d4cdf85cce272a8a687213150339eac50e9e6430ca265dbcef84cd2277e3810c
This archive supplied all 35 selected voltage arrays, the fixed channel rule, source metadata, and full source hashes. The new study uses fixed baseline crops and a new independent-control benchmark.
usesdoi:10.18112/openneuro.ds006519.v1.0.0
Source SEEG spectra and metadata for all 35 records; the exact version and CC0 license are recorded in the pinned dataset metadata.
doi:10.1088/1741-2552%2Faa9ee8

Cited by

1 publication

Public catalog entries on Sciscape.

usesWhole-cycle shifts kept null rejection below 5% in circular phase controls from 35 SEEG records with periodic gaps
We reused the 35 fixed baseline crops, source hashes, source metadata, and retrieval code. We tested a new joint score and a shift correction that keeps periodic gaps. The reused records are the same patient sample.