The last 250 samples gave mixed Brier error changes against 500 samples in eight SEEG task state readers with three tapers

We tested whether the last 250 samples could replace a 500-sample SEEG history in eight fixed readers of task state. The public source was OpenNeuro ds004473 version 1.0.2. Each reader used eight bipolar pairs, three bands, and three DPSS tapers. Readers learned from early trials and predicted states in 674 later trials. Mean Brier error changed from 0.177218 to 0.177332. Error rose in five people and fell in three. The mean paired change was +0.000114, with a 95 percent person bootstrap interval from -0.020202 to +0.019802. This result does not establish equivalent performance. In a fixed secondary check, the last half had lower mean error than the first half with DPSS and Hann tapers. The full and last-half windows had the same endpoint. This test does not measure lower device delay or stimulation benefit. The people and test trials were used in prior work. The package preserves exact digital inputs, source versions and range hashes, code, all readers and predictions, three figures, and reproduction instructions. Independent checks covered 112 readers and 56 complete feature arrays. A clean numerical replay matched 26 output files and 1014 arrays.

stereoelectroencephalography · SEEG · signal processing · Brier error · brain-computer interface · DPSS tapers · window length

Admitted 2026-11-09 20:02 UTC3 citing publications

References

usesThree DPSS tapers gave lower mean Brier error than Hann in eight reused SEEG recordings with 500-sample buffers sci:sha256:6b6f409ccedb538d81478030fd7ac7ab8512d1e3bcef6a9af4feebff6c7e2f5e
This study supplied the exact 500-sample digital buffers, source records, spectral helper, model helper, and retrieval code. We kept its full-window rules and tested first and last halves.
usesEight SEEG pairs classify pre-response and rest windows at 70.7% mean balanced accuracy in later trials from eight recordings sci:sha256:c74068384c7c056974c8568aab987cf5a8eee9979a622c22efa1b67a911e77fd
This study fixed the inherited trial eligibility and early contact pair selection. Its choices define all reader inputs in the present test. Its EDF helper also supports inherited retrieval code.
usesA stationary Gaussian variance correction with 5 percent shrinkage reduced SEEG decoding accuracy for 20 percent simulated sample loss in eight recordings sci:sha256:495cd9ff22cc042e1f138d100dbf95110efd76545b22997f0bb2f49092bdf3ae
This study supplied the original digital voltage extraction and pinned source range records in the inherited input chain. Those records enabled fresh source checks and exact input retrieval.
usesdoi:10.18112/openneuro.ds004473.v1.0.2
The study used the recorded SEEG task data from OpenNeuro ds004473 version 1.0.2; the saved dataset metadata state CC0.
doi:10.1088/1741-2552%2Facae0a

Cited by

3 publications

Public catalog entries on Sciscape.

usesSeparate allowed 12-sample errors at the window center changed all 1,348 tested SEEG decisions within a 100 µV RMS limit per pair in eight people
This work supplied the exact 500-sample pair buffers, trial records, source metadata, range records, and inherited EDF helper. All copied input hashes matched. We used its data and made new error bounds.
usesUsing 100 instead of 500 samples for high gamma raised mean Brier error on later trials in eight SEEG readers with fixed pairs, Hann power, and 500 samples for lower bands
This parent supplied the exact digital bipolar windows, fixed trial records, source range records, metadata, and numerical and retrieval helpers. Its all-band history comparison guided the new separate-band history test.
Later response windows gave lower mean Brier error at both fixed rest positions in eight SEEG recordings with three DPSS tapers
[Legacy relation: reproduces] The first-half and last-half two-band feature arrays and later test logits matched 32 person-method comparisons for DPSS and Hann. The joint-position mean Brier change matched −0.037887 for three DPSS tapers. This is a numerical repeat on the same cohort.
usesLater response windows gave lower mean Brier error at both fixed rest positions in eight SEEG recordings with three DPSS tapers
This work supplied the exact 500-sample digital inputs, selection records, source metadata, EDF headers, pinned range hashes, and power, logistic, and retrieval helpers. Its joint first-versus-last window comparison motivated the new separate response and rest comparison. The same people and later test trials are reused.