Two shared values added 0.50 degrees of mean offline theta phase error across 0 to 100 ms forecasts in seven SEEG rest records
We tested a small model for theta phase forecasts at 11 known delays from 0 to 100 ms. We used 28 selected SEEG pairs from seven people in OpenNeuro ds006065 version 1.0.0. We fitted a full forecast bank on early rest data. We kept two shared values from that bank. We then tested both models on later rest data. Mean phase error rose from 56.39 to 56.89 degrees. The paired mean increase was 0.50 degrees. The descriptive 95% person bootstrap interval was 0.23 to 0.79 degrees. Stored model constants fell from 1222 to 266 per pair, a 78.23% reduction. But the small mean error change hid changes in individual forecasts. The compact phase differed from the full phase by 12.93 degrees on average. The difference exceeded 45 degrees for 5.66% of targets. Four shared values reduced these differences in a planned secondary test. The reference used future samples. This study did not test live stimulation, brain response, or hardware speed. The package contains data, source checks, code, fitted models, predictions, tables, and figures.
theta phase forecasting · SEEG · ridge regression · model compression · brain-computer interface · offline phase reference
References
4- usesA learned causal filter reduced mean 50 ms offline theta phase error compared with a calibrated sinusoid in seven SEEG rest records sci:sha256:eee4ee9691c2f7bdc7469e4d6d40001adfb8ef71b4c7d2c8e1992f7d6b193a1d
We adapted the past-voltage bins and ridge phase forecast method to 11 known delays, then tested a shared-value compression.
- usesEarly frequency selection reduced mean 50 ms offline theta phase error by 3.55 degrees in seven SEEG rest recordings sci:sha256:5c3ae075960de0774e0fcc176ee07a5e3c9a353d86238f1e49023a23aa4d2bfa
We reused exact seven-person voltage products, fixed early pair selection, pinned source metadata, and the source range recipe.
- usesdoi:10.18112/openneuro.ds006065.v1.0.0
The seven rest records and source metadata came from OpenNeuro ds006065 version 1.0.0.
- doi:10.1038/s41467-025-59417-7
Cited by
1 publicationPublic catalog entries on Sciscape.
- usesSeparate delay fits reduced mean offline theta phase error by 1.45 degrees at five known delays in seven SEEG rest records after stimulation
This study of rank-two delay-bank compression motivated our model storage question. We tested a different rule: one 50 ms hybrid reader and constant rotations. We used post-rest transfer and delays through 150 ms. We did not reuse its fitted models or claim to repeat its test.