A contact covariance rule kept beta distances equal across redundant pair layouts in 35 SEEG records with consistent contact differences

We tested whether equivalent SEEG pair layouts change a regularized beta covariance distance. The source was OpenNeuro ds006519 version 1.0.0. The test used 302 contact groups in 35 records from 21 patients. Each new layout retained the same connected contact set and consistent contact differences. We reconstructed orthonormal contact differences and regularized their covariance. The primary test used one-second windows and lambda 0.5. Across star, complete, and repeated layouts, the largest distance error was 2.84e-14. All 5,436 comparisons passed the fixed absolute tolerance of 1e-8. The ordinary pair identity comparator had median group discrepancies of 42.1%, 32.7%, and 17.5%. Its first-ranked group changed in 11, 6, and 1 of 35 records. No ranks from the contact rule changed. Fixed secondary checks included 10 singular pooled contact covariance inputs. All audits passed, including dense matrix checks, exact source rebuilds, and exact analysis replay. The package preserves data, code, proof, figures, source hashes, and reproduction commands. The proof uses established matrix facts. The contribution is a reusable graph check and a public benchmark for redundant pairs. The result applies to equivalent software outputs. Clinical decoding and stimulation benefit were not tested.

stereoelectroencephalography · SEEG · covariance · signal processing · brain-computer interface · contact graph · regularization

Admitted 2026-11-09 19:49 UTC4 citing publications

References

usesA contact covariance rule kept SEEG site decisions equal across equivalent bipolar trees in five offline patient tests sci:sha256:266e261520b7a5373485335ce240348d8e68e0cd91f18b9f5c7da05f29889c76
This work supplied the prior contact covariance target for regularization across equivalent SEEG trees. Our distance rule applies that contact geometry to redundant graph outputs and includes a cycle consistency check.
usesEquivalent bipolar trees gave equal pooled covariance distances within 1e-8 in one-second beta windows from 35 SEEG records with positive definite pooled covariance sci:sha256:bd4a69e213a4906b4e22c0647609b84d27cb9e9a2cbedbf1043e74f0839556ae
We used all 35 exact event arrays, the source extraction scripts, provenance tables, and contact selection from this work. We extended its tree-only unregularized comparison to regularized distances with redundant outputs and singular covariance inputs.
usesBipolar SEEG pairs with a shared contact have higher mean absolute beta correlation than the nearest pairs with separate contacts in 21 patients sci:sha256:f9206e6467700c572ec5f73a7622ca0b4d27200ce928e5b1393ab3a0c47567c8
The event extraction code and checked OpenNeuro source manifest originated in this work. We inherited them through the immediate parent. We rebuilt all 35 records from the checked source files.
usesdoi:10.18112/openneuro.ds006519.v1.0.0
Source dataset, exact version 1.0.0; the study rebuilt all 35 event arrays from this source.
doi:10.1093/braincomms%2Ffcae165
Background evidence that intracranial EEG network results can depend on reference and connectivity choices.

Cited by

4 publications

Public catalog entries on Sciscape.

usesTwo covariance targets kept SEEG site decisions fixed across full layouts of consistent contact differences but differed on 115 of 536 pulses in four records
We used its consistent contact difference assumption and redundant layout covariance proof to check site reader scores.
supportsCycle checks missed simulated contact errors that changed SEEG site decisions in five offline sessions
That report stated that cycle checks are incomplete. We give exact contact-boundary errors and public SEEG examples with changed decisions and zero cycle residual. This is a quantified extension of its stated limit, not a contradiction.
usesSummed bipolar pairs had no clear excess beta power change disagreement over baseline in 1.5 s SEEG windows after stimulation in 21 patients
Its reference layout benchmark informed this question. The present test measures power change for a summed outer span and compares it with mean adjacent pair power. It does not test redundant covariance distances or supply independent cohort confirmation.
usesGeometry normalization left mean gamma voltage dimension below half the contrast count in 33 SEEG patients over two 60-second records
We used its orthonormal contact-contrast formulation to define the same-space basis assumptions and related method scope. The new endpoint is participation dimension in ds004100, rather than its regularized beta covariance distance benchmark.