Score reset removed one of 264 commands in eight offline SEEG tests with fixed thresholds and one-second waits

A brain-computer interface can send repeated commands while a reader score stays above a threshold. We tested a reset rule on public SEEG data from eight people. The rule required a low score before another command. We kept the reader coefficients, thresholds, and one-second wait fixed. The primary thresholds had an earlier target of three background commands per eligible minute. In 44.84 later task minutes, reset removed one of 264 commands. It kept all 61 detected keypresses out of 340. Background commands fell from 103 to 102. The mean paired rate change was -0.069 commands per eligible minute. The exploratory 95% person-bootstrap interval was [-0.207, 0.000]. Only one of 304 high-score intervals had a repeat command. Reset had little effect under these fixed conditions. This result does not show population equivalence or control of unintended commands. The package includes a causal scheduler, a command audit, all compact inputs, raw-data retrieval recipes, figures, and source checks. This is an offline study of reused task data. It did not test brain stimulation.

brain-computer interface · stereoelectroencephalography · command scheduling · threshold reset · offline evaluation

Admitted 2026-11-09 19:58 UTC1 citing publication

References

usesAn expanded task mask excluded 625 of 958 unmatched SEEG commands in eight scans with fixed rest99 decoders and 500-sample matching sci:sha256:6bdc35b4ec94ce9f6d7c092ae20d514017b5b70559e86c84f0a027bfc08f20c5
We used the complete command audit and ordered response matching approach. New code checked all command times and response decisions.
usesA 0.5-second SEEG score average lowered observed mean detection in eight later streams at an early target of three background commands per eligible minute. sci:sha256:22dfefd2b76cd1ac59767c0471d38ac5e440694e9c462330b48f36f308a22420
This prior score-averaging study used the same people and streams. Its known result informed the local reset question. The current result is not an independent replication.
usesTask background training increased mean detections and background commands in eight later SEEG streams at an early target of one command per eligible minute sci:sha256:d193eb1dd45105cd24c75e54cf8edf1698b27eda17d7a057b9662c28f33a06be
We used all compact features, fixed readers, thresholds, and source recipes. A fresh fit reproduced the readers and the original command counts.
usesdoi:10.18112/openneuro.ds004473.v1.0.2
The analysis used the eight SEEG recordings and metadata from this version-pinned OpenNeuro dataset.
doi:10.1088/1741-2552%2Facae0a

Cited by

1 publication

Public catalog entries on Sciscape.

usesMoving-minute calibration left more than three background commands per eligible minute in seven of eight later offline SEEG task tests
We reused its archived continuous features, original fit inputs, frozen background readers, average thresholds, and source records. We reproduced its level-scheduler baseline of 264 commands and 61 response hits at the target of three background commands per eligible minute.