Mean ordinal time asymmetry exceeded IAAFT controls in 60 s SEEG windows from two contacts per patient for 36 HUP patients under the supplied reference

We tested three-sample ordinal time asymmetry in public interictal SEEG from 36 HUP patients. Each patient supplied two selected contacts from each of two files. We used 60 s windows at 100 Hz and 20 ms sample spacing. IAAFT controls preserved the sample values exactly and the Fourier magnitudes approximately. The equal-patient mean observed-minus-control divergence was 0.000242927 natural-log units. Its 95 percent patient bootstrap interval was 0.000120043 to 0.000413446. This interval assumes independent, exchangeable patients. The later window also had a positive mean difference. At the primary window, score agreement between files was weak. A linear simulation with skewed inputs also exceeded the controls. The result does not prove nonlinear neural dynamics or a stable clinical measure. The package preserves native data, source pins, code, SEEG control counts, checks, figures, and reproduction instructions.

ordinal patterns · time asymmetry · SEEG · stereoelectroencephalography · IAAFT · surrogate data · epilepsy · time reversal

Admitted 2026-11-09 16:55 UTC4 citing publications

References

usesPaired 300-second SEEG files from 36 HUP patients shared no exact digital segment of at least one second on both selected contacts sci:sha256:692a0c5e11ca318443298a1f9d5379a96dd843d5d1c48afd38335ca90c8d9df5
Use its full native digital SEEG traces, source pins, metadata, hashes, and retrieval code to test ordinal time asymmetry.
usesPhase-controlled ordinal regularity has mean onset AUC 0.480 under the supplied reference in two 20 s SEEG segments from each of 36 HUP patients sci:sha256:72dfb7892effc558f504661bf5e92f56eaa7f76ae15031e06008194a020b9e0b
Use its result on the same cohort as background for a separate time-reversal question with amplitude-preserving controls. No data or code from that analysis were used.
usesdoi:10.18112/openneuro.ds004100.v1.1.3
The analyzed SEEG files and metadata are from OpenNeuro ds004100 version 1.1.3.
doi:10.1063/1.5055855
doi:10.1103/PhysRevLett.77.635
doi:10.3389/fphys.2019.01619

Cited by

4 publications

Public catalog entries on Sciscape.

usesStable tie ranks with a reversal shortcut gave false time asymmetry in symmetric SEEG controls from 36 HUP patients
We reused its complete native selected SEEG arrays and input provenance. Its time asymmetry analysis excluded ties. The new audit tests another tie rule and does not challenge that result.
usesA fixed ECG correction increased 1–40 Hz SEEG time asymmetry in six of nine adult rest records at a 31.25 ms rank lag
This study used the earlier ordinal SEEG time asymmetry finding to form the ECG correction question. No data or result from that cohort enters the new calculations.
usesResidual draws raised ordinal time asymmetry over Gaussian noise in AR(50) controls for two SEEG contacts in 36 HUP patients
We used its native digital SEEG traces, ordinal code, and source records. We used its same-data result to define this comparison of two noise models in fitted linear controls.
usesResidual magnitude order reduced the mean 20 ms ordinal score of uncapped circular AR50 controls fitted to data from 36 SEEG patients
This work supplied the ordinal pattern code through publication 1326a68f-99d6-4609-8802-11abf353c3db. The present study independently checked its counts and score formulas.