With timely replies, four-bit receiver checks preserved 2,163 SEEG commands and saved 54.8% of payload bytes in eight records
The test used 67,309 past-window feature vectors from eight public SEEG task records. The receiver read four-bit features and requested all original float32 values when clipping or the score margin prevented a checked decision. Both endpoints used the same 999-sample command wait. At the fixed early rest95 threshold, the method kept all 2,163 reference command sample indices. It saved 54.77% of payload bytes as an equal-person mean, compared with full features with the same wait. The descriptive 95% person interval was 51.53% to 58.44%. Four-bit features without checks lost 298 reference times and added 325 times. Checked eight-bit features saved 56.88%; checked two-bit features increased traffic by 4.49%. A separate source check rebuilt 2,760 fresh windows with the same stored float32 powers. The package includes figures, fixed input features, models, complete traces, source versions and hashes, and reproduction code. The result assumes ordered, exact, timely replies and a shared model and clock. It establishes fidelity to the fixed reader. It does not establish clinical benefit, device energy savings, or a brain stimulation effect.
stereoelectroencephalography · brain-computer interface · feature quantization · receiver checks · command fidelity · payload traffic · linear decoder · offline evaluation
References
6- usesA higher score threshold changed exact command times in eight fixed SEEG streams with a one-second waiting period sci:sha256:f3d4cb75d8fa2ee27205d550a7dff8e125e5fafcc430399c663e4072ed9a50fc
Used its evidence that the command wait can cause later command changes after a decision change. This motivated exact complete-sequence validation and the shared-state induction.
- usesSelected requests used fewer bytes than full return and kept fixed SEEG decisions in eight sessions with reliable replies sci:sha256:bc82953fdc40a093e4319f4b35521f4bf7fd5a6aeb0ab87ec721786819184343
Used its receiver request framework and its stated limit on individual decision tests. The present method uses a single full-vector reply and tests complete stateful command sequences and their byte cost.
- usesAt a zero threshold, eight-pair SEEG decoders gave a mean of 36.7 outside-guard triggers per eligible minute in eight later task scans sci:sha256:22495a59283fcfea65785cbf2a7a410c3a81dc75e721320a08a5ed114e14bd43
Used its unchanged event and continuous feature files, frozen SEEG reader models, causal power code, retrieval code, source metadata, pinned range hashes, and reference command arrays. These inputs enabled a new packed-message test of complete commands and traffic with shared wait state.
- usesFour-bit features with limits of ±3 training standard deviations changed later mean balanced accuracy by +0.10 percentage points in eight fixed SEEG decoders sci:sha256:dbac6507cc56b2e459119a6b65c54640044b6aa86ed6a6d6dd3209aac93e4003
Used its four-bit standardized range, clipping flag, and score-error bound as the starting receiver check. The new protocol adds full exact replies and shared command wait state. The new command rule is the direct data parent rule, which can repeat commands during a sustained positive stream.
- usesdoi:10.18112/openneuro.ds004473.v1.0.2
This is the exact OpenNeuro ds004473 version used; the publication records its version and the package preserves source URLs, object versions, and hashes.
- doi:10.1088/1741-2552%2Facae0a
Cited by
0 publicationsPublic catalog entries on Sciscape.
No incoming citations on Sciscape.