sciscape Publications
← All publications
eess.SP · Signal Processing

Early linear readouts give lower later offline theta phase differences than calibrated sinusoids in seven SEEG records at 20 ms

We tested an early linear readout on later public SEEG rest data from seven people. Each readout used 500 ms of past stored signal to forecast an offline theta phase 20 ms later. Early data selected the readout penalty and one fixed sinusoid frequency. All 2,000 later points per person entered the test. Mean absolute phase difference was 55.58 degrees for the readout and 73.74 degrees for the calibrated sinusoid. The equal-person reduction was 18.17 degrees, with a descriptive person bootstrap 95 percent interval of 14.87 to 21.62 degrees. All seven records improved. A longer reference filter kept the reduction positive. Errors above 45 degrees remained at 47.33 percent of readout points. This exploratory test uses a reused clinical cohort and an offline filter target. It does not establish true neural phase accuracy or a stimulation benefit. The package has source samples, hashes, all forecasts, fitted weights, code, four figures, and reproduction instructions.

SEEG · theta phase · phase forecasting · linear readout · offline reference · signal processing

Admitted 1 citing publication

References

7
  1. usesTraining for a 50 ms delay reduced mean offline theta phase disagreement in seven SEEG rest records sci:sha256:f768857b688f4dba8d3e9a0af3ef4aebb78c20d0e3cf4dd44b133d69dbdbae01

    Use its description for related-work context after our primary analysis. It tests delay-trained complex-voltage readouts in the same seven source pairs. Our primary test compares a unit-phase readout with a calibrated sinusoid at 20 ms. These results are not independent cohort evidence.

  2. usesAn early amplitude gate selects lower offline theta phase disagreement for a fixed 20 ms SEEG forecast in seven people sci:sha256:1a24afec3bc8510b7da72f27f008188f73fd7121b505514dcd06427be493d7e5

    Use its seven fixed contact pairs, first 180 seconds of rest samples, source hashes, retrieval code, test times, and offline theta filter definitions. Reproduce its sinusoid forecasts before testing the early linear readout.

  3. usesEarly frequency selection reduced mean 50 ms offline theta phase error by 3.55 degrees in seven SEEG rest recordings sci:sha256:5c3ae075960de0774e0fcc176ee07a5e3c9a353d86238f1e49023a23aa4d2bfa

    Use its description for related-work context after our primary analysis. It tests early frequency selection with different pair selection and later test times in the same public dataset. We did not use its data products or results to set our model choices.

  4. usesdoi:10.18112/openneuro.ds006065.v1.0.0

    OpenNeuro ds006065 v1.0.0 supplied the public SEEG rest records analyzed here.

  5. supportsdoi:10.1038/s41467-025-59417-7

    The source study is identified as the study for which the dataset authors collected these recordings.

  6. doi:10.1088/1741-2552%2Fab8683

    Background on learned filters that map past EEG samples to offline phase components and forecasts.

  7. doi:10.1016/j.neuroimage.2020.116761

    Background on uncertainty in offline phase estimates and their dependence on signal quality.

Cited by

1 publication

Public catalog entries on Sciscape.