Division by 2–40 Hz power increased baseline and post-stimulation high-gamma feature drift in two SEEG datasets
We tested whether division by 2–40 Hz power made high-gamma features more stable after brain stimulation. We used public SEEG spectra from 103 visual records in 19 people and 35 motor records in 21 people. The datasets were tested separately. The ratio had more raw drift in every person in each dataset, both within baseline and from baseline to the marked post-stimulation period. The median paired increase in post-stimulation drift was 1.354 dB in the visual dataset and 1.148 dB in the motor dataset. The primary visual test did not detect a reduction in excess drift. The median reduction was −0.050 dB, with a 95% bootstrap interval from −0.184 to +0.061 dB. The exact sign test gave p = 0.648. This result applies to two-second windows and the tested power ratio. It does not measure BCI decoder accuracy. The package preserves spectra, code, results, three figures, a fixed local plan, and raw data retrieval instructions.
stereoelectroencephalography · high-gamma power · feature drift · electrical stimulation · power ratio · brain-computer interface · signal processing
References
6- usessci:sha256:3d1647c9a767ea0b389c4475596fc0a13c65b2f5d64e1cb07e481cdea7a95f9f not known to Sciscape
This study supplied all 112 visual SEEG spectra, source metadata, raw file hashes, and original extraction code. We used them for the new paired ratio test. We reproduced its absolute-power person scores.
- usesA paired test did not detect increased high-gamma drift in annotated post-stimulation windows from 21 SEEG participants sci:sha256:0eb1cda13cce113a7a1fe70466917eeb0278c662b2155ce27ee9d36ce7636489
This study supplied all 35 motor SEEG spectra, source metadata, raw file hashes, and original extraction code. We used them for the new paired ratio test. We reproduced its absolute-power person scores.
- usesdoi:10.18112/openneuro.ds005169.v1.0.0
Source spectra and metadata for the visual stimulation dataset.
- usesdoi:10.18112/openneuro.ds006519.v1.0.0
Source spectra and metadata for the negative motor response dataset.
- doi:10.1002/hbm.25749
- doi:10.1016/j.neuroimage.2026.121822
Cited by
2 publicationsPublic catalog entries on Sciscape.
- challengesChannels kept by an 80% baseline rule had lower rates of high-gamma change above 3 dB in separate later windows of 99 visual and 33 motor SEEG records
The earlier report states that the sources do not provide a link between person identities across releases. The public participant tables define seeg_id as a source patient ID. One ID links visual sub-11 with motor sub-13, with matching contact names and coordinates. This corrects that metadata claim under the assumption that source IDs are unique and consistent. It does not challenge the earlier power-ratio calculations.
- usesChannels kept by an 80% baseline rule had lower rates of high-gamma change above 3 dB in separate later windows of 99 visual and 33 motor SEEG records
This package supplied the preserved visual spectra, both source manifests and metadata, and earlier motor spectra used for the fresh source comparison. The new test uses a separate later-window outcome.
- usesA fixed beta SEEG model had person median error above zero output in both test periods after simulated shaft loss
Its marked-window limits and distinction between feature drift and decoder accuracy informed the scope. The present test uses fitted beta-voltage reconstruction, not its high-gamma ratio or an independent cohort.