Refitted spectrum controls gave 20.82 percent forecast skill at 50 ms for a future theta filter term in seven SEEG rest records
We tested a voltage forecast that can help assess brain-interface timing methods. The data were 28 selected SEEG pairs in seven public rest records from OpenNeuro ds006065 version 1.0.0. The target was the part of a centered theta filter that used voltage at or after the input cutoff. At a 50 ms target-center horizon, mean squared-error skill was 20.62 percent in recorded voltage. The mean was 20.82 percent in 99 refitted spectrum controls per person. The recorded-minus-control difference was -0.19 percentage points. Its descriptive 95 percent person bootstrap interval was -1.35 to 1.00 points. The controls kept each block spectrum and all cross spectra. They changed Fourier phases. A shorter-block check gave a difference of -0.37 points. These results show that this benchmark can give positive skill after phase changes. They do not establish equal performance, biological phase, or stimulation benefit. This was an exploratory use of seven reused clinical records. The package includes exact input data, source checks, all model weights, full score tables, code, and four figures.
stereoelectroencephalography · voltage forecasting · Fourier phase controls · spectrum preservation · brain-computer interfaces · ridge regression · theta filter
References
6- A 500 ms past voltage sum gave a 63.26 degree error for centered theta phase 50 ms ahead in seven SEEG rest records sci:sha256:678ac0e1b4e8d00b5e72eb22a8e9bbbfc9c4bd67c224a77925cb4cba677fc26f
[Legacy relation: reproduces] We use its exact seven input voltage records and source metadata. We reproduce all four unavailable theta filter-term forecast scores. We then compare those scores with refitted spectrum controls.
- usesLearning reduced mean 50 ms offline theta phase error by 11.25 degrees in spectrum controls from seven SEEG rest recordings sci:sha256:be46eec0bf0062fd4ea3fded570f754676cc2c3a9e4df6be657676ba2ae8d0c5
We use its shared Fourier-phase control and model refit method as context. The present targets exclude all pre-cutoff voltage. We use new later test blocks and a shorter-block check.
- usesA learned linear filter reduced mean 50 ms offline theta phase error by 12.54 degrees in seven SEEG rest recordings sci:sha256:681793db790925fefcdd246f20a125cf7ec0a4d0c7ac415b89cd84c8e64b5d35
Its long-window bipolar voltage package is the earlier source of the exact data used through the direct input publication. Its ridge objective supplies the reader design.
- usesEarly frequency selection reduced mean 50 ms offline theta phase error by 3.55 degrees in seven SEEG rest recordings sci:sha256:5c3ae075960de0774e0fcc176ee07a5e3c9a353d86238f1e49023a23aa4d2bfa
We keep its early selection of four disjoint adjacent SEEG contact pairs per person. No new contact selection enters this study.
- usesdoi:10.18112/openneuro.ds006065.v1.0.0
Source dataset for the seven pre-stimulation rest records and 28 inherited bipolar SEEG pairs.
- doi:10.1038/s41467-025-59417-7
Background paper for the hippocampal network stimulation study associated with the source dataset.
Cited by
4 publicationsPublic catalog entries on Sciscape.
- usesCommon spectrum controls gave a 0.22 percent error decrease from additional pairs for linear forecasts of future theta terms in seven SEEG rest records
Used exact four-pair voltage inputs, source metadata, pinned source objects, and the common-phase control method. Tested additional pairs with full model choice and fitting on every control.
- usesSpectrum controls gave 50.75 percent mean skill and a causal 7 Hz reader gave 50.41 percent at 50 ms in nine SEEG rest records
We adapt its refitted block Fourier phase control design and short-block check. It tested a centered future-filter target in seven records. We test a causal state with only 50 ms of unseen voltage in nine records.
- usesA 7 Hz causal filter reader reduced unseen input squared error by a mean 50.4 percent at 50 ms in nine SEEG rest records
Its spectrum controls guided interpretation of positive linear prediction skill. We make no claim that the new causal-state result exceeds what recorded temporal covariance can explain.
- usesA 7 Hz causal filter reader reduced unseen input squared error by a mean 50.4 percent at 50 ms in nine SEEG rest records
Its spectrum controls guided interpretation of positive linear prediction skill. We make no claim that the new causal-state result exceeds what recorded temporal covariance can explain.