sciscape Publications
← All publications
q-bio.NC · Neurons and Cognition

SEEG power did not show a clear mean gain over task inputs for 100 ms keypress forecasts during the Go Cue state in eight people

We tested 100 ms keypress forecasts during the observed Go Cue state in eight public SEEG recordings. All eligible stored cue times entered the chronological fit, tune, or test split. Future responses set labels but did not select times. On 2,710 test windows, the equal-person Brier score was 0.107130 for task inputs and 0.105191 with SEEG. The paired difference was -0.001939. Its descriptive 95 percent person interval was -0.005990 to +0.001298. Five people improved and three did not. Without person 11, the mean difference was -0.000232. The evidence did not show a clear mean gain. It did not show a zero effect. The package provides code, input power, source metadata, fitted weights, every forecast, four figures, and reproduction instructions. Independent labels, an independent optimizer, and fresh pinned raw ranges passed checks. This is an exploratory benchmark from reused clinical task data. No live BCI or stimulation was tested.

stereoelectroencephalography · SEEG · brain-computer interface · keypress forecasting · Brier score · task-state decoding

Admitted 0 citing publications

References

6
  1. usesSEEG power readers gave no clear error reduction in 134 later cue-matched 100 ms keypress pairs from eight patients sci:sha256:8677d3b0ad27a0ccffbc5096601cb48dc45940a4e7ef047a7bb58e06a4c7785c

    Its matched cue-state study informed the question and task control. Our test uses every eligible stored cue endpoint at its observed event rate. It does not match cases or select times from future responses.

  2. usesFixed SEEG power models gave no clear gain beyond task events for continuous 100 ms keypress forecasts in eight people sci:sha256:5a550bbec827507f5125afef09901460929b3204e989fb4167f8ca343bdf5140

    We used its continuous SEEG power, source metadata, task input functions, model solver, and frozen all-time models. We fit new models only during the observed Go Cue state and scored the same later cue endpoints.

  3. usesSEEG key decoding from the first fast block had a mean AUC of 0.54 in later slow blocks in six patients sci:sha256:6407f4f45a2db401fef737a5f946c2656d5af25ba66dbcab30e4417db2c4fc72

    Its timing concern for person 11 informed the planned summary without that person. This check did not replace the main eight-person result.

  4. usesShared scores gave no clear Brier error gain for 100–500 ms keypress forecasts in eight SEEG recordings with known cues sci:sha256:e4b3b21c9bed9bb0b9c7fb7a02623692dcd5ecac843f7b2a28c65defcb90b02b

    Its multi-horizon study informed the scope comparison. It selected complete retained trials. Our 100 ms benchmark uses observed cue state to select endpoints before any future response is used as a label.

  5. usesdoi:10.18112/openneuro.ds004473.v1.0.2

    Source dataset used for the eight analyzed SEEG records; the publication identifies OpenNeuro ds004473 version 1.0.2.

  6. doi:10.1088/1741-2552%2Facae0a

Cited by

0 publications

Public catalog entries on Sciscape.

No incoming citations on Sciscape.