Adding early delay SEEG burst features raised mean loss in a fixed recall model for 13 patients under source trial order
We tested whether early delay burst features improve later recall forecasts in public SEEG data from 13 patients and 14 sessions. We used thresholds from training data and a causal beta and gamma detector. We added burst counts and active fractions to a fixed power and task context model. The test had 751 trials. Mean patient Brier loss changed from 0.179663 to 0.192827. The gain was -0.013164, with a 95% patient resampling interval of [-0.025688, -0.003348]. The interval was below zero. The tested feature set increased forecast loss in this cohort. This detector found few sustained events. The released data were already cleaned and selected. This is an offline test under source trial order. It does not establish live decoder or stimulation performance. The package preserves code, features, forecasts, figures, source hashes, and a full data retrieval recipe.
SEEG · working memory · recall forecasting · burst features · Brier loss · causal signal processing · OpenNeuro
References
3- usesEarly delay SEEG power gave no mean Brier gain in later recall forecasts for 13 patients under source trial order sci:sha256:adbcbe06fcce87132c9633b5137a24af646e5c49afc3ac2d09469a1e05618474
We used its EDF reader, retrieval code, source manifest, matched power features, chronological split, and fixed power and task context model. We checked rebuilt power features and baseline predictions against its preserved files. These assets enabled a fixed comparison that adds causal burst features on the same cohort and split.
- usesdoi:10.18112/openneuro.ds006136.v1.0.1
Source SEEG epochs, channel metadata, original trial IDs, and recall labels from OpenNeuro ds006136 version 1.0.1.
- doi:10.1016/j.neuroimage.2026.121718
Cited by
0 publicationsPublic catalog entries on Sciscape.
No incoming citations on Sciscape.