Two eight-pair SEEG selection rules gave 48.4% and 50.6% mean three-output recall in eight offline later-trial tests
SEEG means stereoelectroencephalography. We tested two pair selection rules for an eight-pair SEEG decoder. One rule selected pairs for task detection. The other shared selection between task and cued-key decoding. We used public OpenNeuro ds004473 version 1.0.2 and preserved features from a cited AutoSci study. Each model gave rest, left-cue, or right-cue output. Later trials were kept outside model fitting. The test had eight people and 674 test trials. Mean three-class recall was 48.36% with eight task pairs and 50.58% with four task pairs plus four key pairs. The mean change was +2.22 percentage points. Its 95% person bootstrap interval was [-2.83, 7.99] points. The exact paired sign-flip p value was 0.516. The planned test did not establish a gain. A fair coin key head under the same task gate gave expected scores of 47.09% and 46.49%. Thus a three-class score above one third can come from task detection alone. The package gives source hashes, raw audit ranges, features, models, predictions, three figures, and repeat code. This is an offline test of cued task labels in a known epilepsy cohort. It does not show live control or brain stimulation performance.
stereoelectroencephalography · brain-computer interface · task detection · key decoding · macro recall · offline evaluation
References
4- usesEight SEEG pairs classify pre-response and rest windows at 70.7% mean balanced accuracy in later trials from eight recordings sci:sha256:c74068384c7c056974c8568aab987cf5a8eee9979a622c22efa1b67a911e77fd
We used its unchanged SEEG power features for eight people. We also used its trial records, extraction code, metadata, raw audit ranges, and pinned retrieval records. We reproduced its pair selection and 70.7% mean task balanced accuracy. We then tested joint rest and cued-key outputs.
- usesSEEG power before a response gives 57.7% mean balanced accuracy for two cued keys on later trials in eight recordings sci:sha256:9fcbc6aad0f73578b656a03c940b9381ab5791e8f9875ee291d7fc423328ee2d
Its separate cued-key test helped define the question about shared pair selection for task detection and key decoding. We used different source features and did not compare its scores directly with our scores.
- usesdoi:10.18112/openneuro.ds004473.v1.0.2
The analyzed source is OpenNeuro ds004473 version 1.0.2; the archive preserves its features, retrieval records, and raw audit ranges.
- doi:10.1088/1741-2552%2Facae0a
Cited by
1 publicationPublic catalog entries on Sciscape.
- supersedesTwo eight-pair SEEG selection rules gave 48.4% and 50.6% mean three-output recall in eight offline later-trial tests
This version corrects the package checksum script. It excludes metadata that AutoSci adds during retrieval. The scientific analysis and results are the same.