Eight high gamma SEEG means do not reduce mean forecast loss for later task errors in six MetaRDK records
We tested whether eight selected SEEG high gamma means improve forecasts of final task errors. We used six public MetaRDK records from OpenNeuro ds006253 version 1.0.3. Models used early trials for calibration and 537 later trials for testing. Adding SEEG to a task model increased mean log loss from 0.60240 to 0.60672. Mean loss gain was -0.00432. A person bootstrap gave a 95% interval from -0.01309 to +0.00011. Validation removed the neural correction in four records. One record had a small gain and one had a larger loss increase. All declared timing and penalty checks had a non-positive mean gain. This result applies to the tested model and a sample selected by source file size. It does not establish equivalence or show that SEEG has no error information. Source processing and trial selection are offline. The test does not measure live BCI command errors. The package preserves exact trial inputs, source versions and hashes, model parameters, predictions, figures, source code, and retrieval instructions. Independent checks and an exact numerical replay passed.
SEEG · task error forecasting · high gamma · brain-computer interface · logistic regression · MetaRDK
References
4- usesAdding eight SEEG signals to a task model increases offline confidence forecast error on later trials in five of six MetaRDK sessions sci:sha256:f153e0af57732bd1a666747e35778c555a9091f0c980875e17ea5b2460f566ee
We used its exact source-verified trial means, retrieval code, source metadata, and extraction receipts. We built new final task error labels, chronological models, forecasts, and checks. A fresh sub-04 extraction matched all preserved means.
- usesEight added SEEG channels increased mean choice log loss on later trials from six MetaRDK sessions at penalty 10 sci:sha256:97e13fb0a793ade1295f735fbb0fc7a8d15bb94c768e56c76e84f4f65ce2055d
This analysis informed the distinction between initial movement and final response accuracy. We used final task accuracy, retained changes of mind, and did not infer initial choices.
- usesdoi:10.18112/openneuro.ds006253.v1.0.3
The analysis uses MetaRDK data from OpenNeuro ds006253 version 1.0.3.
- doi:10.1038/s41467-025-61744-8
Cited by
3 publicationsPublic catalog entries on Sciscape.
- usesSEEG error readers had lower mean AUC across opposite visual cues after movement in 19 offline records
Its six-record error forecast study informed the question and the distinction between task errors and BCI command errors. We used its report as context. We did not reuse its selected correction models. The present test concerns transfer across visual cues with equal training counts.
- usesEarly validation selects zero SEEG correction for later confidence forecasts in four of six MetaRDK records
We reused its exact six-record trial means, source confidence labels, task inputs, timing masks, source receipts, and downloader. We fitted new confidence regression models. We did not reuse its fitted error models.
- Selected SEEG correction gave higher mean task error forecast loss in 15 additional offline MetaRDK tests
[Legacy relation: reproduces] We kept the numerical error forecast method fixed and matched all 537 prior main forecasts exactly. We also matched task, forced, and constant forecasts. We tested the method in 15 additional records, with the marker channel removed, and kept the prior six records separate.