Output shifts gave excess beta PSI rejections for shared contacts under independent contact segment rotations in two SEEG records
We tested phase slope index (PSI) shift controls with two public resting SEEG records. The null independently rotated the sequence of recorded two-second segments for each contact. Each rotation kept all frequency bins together. We used 28 contact quartets and 400 seconds per record. Each quartet had 2,000 simulated observations. For shared-contact bipolar pairs, output shifts gave 18.01% rejections at a 5% threshold. Contact shifts gave 4.81%. These rates have equal patient weights. The paired difference was 13.20 percentage points. Its simulation standard error was 0.13 points. Disjoint pairs gave 4.89% output rejections. We proved that output shifts give a valid rank test for disjoint pairs under this null. The proof fails in general for shared pairs. The first 200 seconds also showed excess shared-pair rejections. The package preserves complex inputs, offsets, null scores, counts, code, source locks, proofs, and reproduction checks. The result applies to this conditional rotation model and these two records.
phase slope index · signal processing · SEEG · surrogate tests · shared contacts · segment rotations
References
4- usesOutput phase tests gave excess beta phase slope index flags under independent contact phases in two public SEEG spectral models sci:sha256:a3d9910fc20f1ef9e604d7af98ee46359e232cf4c0f6a90e066f3096d94c2046
We used the quartet selection rule, native EDF reader, source locks, BIDS metadata, and small reader fixtures. We downloaded the raw files again and extracted complex coefficients. We tested recorded contact segment rotations instead of independent phases in each frequency bin. The study uses the same two patients.
- usesSignal shift order changed correlation nulls for pairs with a shared contact in 36 public SEEG recordings sci:sha256:536b16601141a53bc122d960c0acfb4876559090fa2a4f719a0afc5caa4e0268
The shared-contact correlation result supplied the shift-order question. We tested that question with raw beta PSI and recorded segment rotations. This publication supplied no signal data to the new analysis.
- usesdoi:10.18112/openneuro.ds005670.v1.0.0
The analysis uses OpenNeuro ds005670 version 1.0.0; the publication records the exact version and source hashes.
- doi:10.1103/PhysRevLett.100.234101
Cited by
1 publicationPublic catalog entries on Sciscape.
- usesFull contact ranks controlled beta PSI test size under independent uniform segment rotations in two SEEG records
We used the fixed 28-case selection, complex source coefficients, raw object lock, source metadata, EDF reader, retrieval code, independent decoder code, and prior calibration evidence. We add complete relative-contact ranks, a common-rotation proof, a reusable exact test, and an exact output-shift benchmark on the same records. This is a secondary method analysis, not independent patient evidence.