Early products showed no clear gain over squared means for ten beta SEEG pairs per record in 21 patients
We tested two rules that select ten SEEG channel pairs from two early beta correlation windows. The product rule used a*b. The control used the square of the mean of a and b. We used all 35 recordings from 21 patients in OpenNeuro ds006519 version 1.0.0. Two later windows gave mean correlation products of 0.14226 and 0.14292 for the selected pairs. The patient mean difference was -0.00065. The descriptive 95% bootstrap interval was [-0.00769, 0.00581]. The exact two-sided sign-flip p value was 0.8994 under independent patients and symmetric null differences. The rules shared 92.3% of their selected pairs on average. This test showed no clear gain from the early product rule. It did not show equivalence or improved BCI performance. Simulations showed why low average score bias does not ensure better selection. The package contains code, all event arrays, all primary pair values, selection records, four figures, and a fixed source retrieval recipe. All 355 source files passed hash and version checks. A raw-data rebuild matched all 35 event arrays exactly.
stereoelectroencephalography · SEEG · brain-computer interfaces · correlation · pair selection · selection bias · reproducibility
References
3- usesBeta power ranks have higher agreement than beta network strength ranks between 3 s baselines in eight SEEG patients sci:sha256:dd8a1ee586c1a7276d197e8b391188542f3b8a9603e2240b0f58a49961c495b7
Its reported variation in beta network ranks across repeated patients helped motivate a test of selection on later estimates. We did not treat its reported results as new observations.
- usesA baseline rule reduced later beta SEEG correlation error in 1 s windows from 21 patients sci:sha256:05572bbfd69c0cd02a49e72246e6e438e50bf8fda0ee17b297b3702b9d6fccab
We used its 35 event arrays, extraction code, contact parser, source manifest, and raw retrieval recipe. We checked all imported files and rebuilt the arrays from source files. Its earlier prediction-error study informed this new later-target selection test.
- usesdoi:10.18112/openneuro.ds006519.v1.0.0
We analyzed all 35 recordings in OpenNeuro ds006519 version 1.0.0. The saved dataset metadata identifies this exact release and declares CC0.
Cited by
1 publicationPublic catalog entries on Sciscape.
- usesLocal 2 s beta baselines gave no clear gain over transferred baselines for post-stimulation correlation forecasts in eight SEEG patients
We use its exact 22 event arrays, reviewed extraction and retrieval code, contact parser, and source identity checks. We rebuilt the arrays from the public source and checked every field.